Non-steroidal anti-inflammatory drugs in prevention of gastric cancer

Abstract

Non-steroidal anti-inflammatory drugs (NSAIDs) including cyclooxygenase 2 (COX-2) selective inhibitors, are potential agents for the chemoprevention of gastric cancer. Epidemiological and experimental studies have shown that NSAID use is associated with a reduced risk of gastric cancer although many questions remain unanswered such as the optimal dose and duration of treatment. The possible mechanisms for the suppressor effect of NSAIDs on carcinogenesis are the ability to induce apoptosis in epithelial cells and regulation of angiogenesis. Both COX-dependent and COX-independent pathways have a role in the biological activity of NSAIDs. Knowledge of how NSAIDs prevent neoplastic growth will greatly aid the design of better chemopreventive drugs and novel treatments for gastric cancer.

Key Words: Nonsteroidal anti-inflammatory drugs, Gastric cancer, Cyclooxygenase, Prevention, Intervention

INTRODUCTION

Gastric cancer is a major cause of death in the world. However, the etiology of gastric cancer is not fully understood. Although early diagnosis and treatment of gastric cancer significantly improve prognosis[1,2], the 5-year survival rate is only 10%-15% in those with advanced disease[3]. Therefore, the primary prevention of gastric cancer is of particular importance. Several studies have shown that human gastric carcinoma expresses significantly higher levels of cyclooxygenase (COX)-2 than non-cancerous adjacent mucosa[4], suggesting that COX-2 plays a possible role in gastric carcinogenesis. Numerous observations and clinical studies have suggested that the use of NSAIDs is associated with reduced incidence of gastric cancer[5-8]. NSAIDs, particularly the highly selective COX-2 inhibitors, are promising chemopreventive agents[9-12].

Although the effects of NSAIDs on gastric cancer remain unclear, both COX-dependent and independent mechanisms may have roles in the biological activity of NSAIDs. In this review, we concentrate on the use of NSAIDs for gastric cancer prevention and the underlying mechanisms. Furthermore, we also suggest research opportunities on clinical application of NSAIDs for gastric cancer prevention.

NSAIDS AND GASTRIC CANCER

Epidemiological evidence

There are indications, although still limited and preliminary, of a protective effect of NSAIDs on gastric cancers. The risk of gastric cancer is reduced, though not significantly, in the Swedish[13] cohorts of patients with rheumatoid arthritis, with the relative risk (RR) of 0.63 (95%CI = 0.5 to 0.9) for NSAID users. In the American Cancer Society cohort, there is a significant reduction in the risk of death for gastric cancer in aspirin users (RR = 0.64, 95%CI = 0.51 to 0.80). In addition, a dose-dependent trend in risk reduction has been observed, with the RR of 0.73 (95%CI = 0.56 to 0.96) for occasional use (1-15times per month) and 0.49 (95%CI = 0.33 to 0.74) for use 16 or more times per month[5]. With regard to case-control studies, in an American study of cancers of the esophagus and stomach, based on 261 adenocarcinomas of the cardia, 368 noncardia and 695 population controls, current aspirin users have a lower risk for noncardia cancer (RR = 0.46, 95%CI = 0.31 to 0.68), but not for cardia cancer (RR = 0.80, 95%CI = 0.54 to 1.19) when compared to never users. However, there is no significant trend for the risk reduction of gastric cancer with the frequency and the duration of aspirin use[6]. A case-control study from Russia was conducted on 448 gastric cancer cases and 610 hospital controls. The use of aspirin is significantly associated with a risk reduction for the development of gastric cancer (RR = 0.49, 95%CI = 0.31 to 0.77), although the favorable effect seems to be limited for patients with noncardia cancer and Helicobacter pylori (H pylori) infection[7]. Another case-control study from Sweden, conducted on 567 cases and 1165 hospital controls, reported that users of aspirin have a moderately reduced risk of gastric cancer compared to never users (RR = 0.7, 95% CI: 0.6 to 1.0), and gastric cancer risk falls with the increasing frequency of aspirin use. Moreover, the risk reduction is statistically significant not only for cardia but also for noncardia cancer, and is slightly more marked in patients with H pylori infection[8].

However, there are several limitations for the above observational studies. First, the small number of cases may result in inadequate statistical power[7,8,15]. Second, the choice of control subjects in case-control studies may distort the results because hospital-based controls can not represent population-based controls[16]. Third, different definition of drug exposure and different doses of drugs have been evaluated in these studies. Therefore, the optimal duration and dose of treatment could not be recommended from these studies.

A recent systematic review with meta-analysis of nine observational studies evaluating the association between NSAID use and the risk of gastric cancer suggested that long-term use of aspirin or other NSAIDs is associated with a statistically significant, dose-dependent reduction in the risk of gastric cancer (OR = 0.78, 95%CI = 0.69 to 0.87). When the analysis is stratified by the site of gastric cancer, use of NSAIDs is associated with a statistically significant lower risk for noncardia gastric cancer (OR = 0.72, 95%CI = 0.58 to 0.89), but not for gastric cancer at the cardia. However, whether the minimum effective duration of NSAID use can reduce gastric cancer risk remains undetermined because of the relatively small number of studies, the use of different cut-off time points of treatment duration, and varying definitions of drug exposure[14]. More prospective studies are needed to clarify whether a duration effect exists in the relationship between NSAID use and gastric cancer risk.

Animal models and in vitro studies

The results obtained from animal models evaluating the effect of NSAIDs on gastric cancer have been conflicting. Two studies on animal models of carcinogens suggested that NSAIDs (aspirin and flurbiprofen) could enhance chemical-induced gastric carcinogenesis in rats[17,18], indicating that prostglandins (PGs) produced by the gastric mucosa protect against the action of carcinogens in this rodent model. In contrast, broad anticarcinogenic effects of NSAIDs have been demonstrated both in vivo using laboratory animals and in vitro using cancer cell lines. Chemically-induced gastric tumors can be reduced by aspirin[19], indomethacin[20], sulindac and ibuprofen[21] in mice. Sawaoka et al[22] have investigated the effects of NS-398, a selective COX-2 inhibitor, or indomethacin, a non-selective COX-2 inhibitor, on the growth of gastric cancer xenografts in vivo. Both drugs could reduce the tumor volume significantly in a dose-dependent manner. Fu et al[23] also found that sulindac and celecoxib obviously decrease the blood vessel quantity and tumor size of SGC7901 cancer xenografts. In a recent study Hu et al[24] showed that treatment with celecoxib could suppress MNNG-induced gastric cancer in rats. Indomethacin, sulindac and NS-398 significantly inhibit the proliferation and growth of human gastric cancer cell line MKN28 in vitro[25,26].

POSSIBLE MECHANISMS OF CANCER PREVENTION

The results described above provide convincing evidence that non-selective NSAIDs and selective COX-2 inhibitors can inhibit gastric cancer cell formation and growth. However, the molecular basis for the striking chemo-preventive effects has not been fully clarified. Several studies indicate that COX-2 expression is a relatively early event during carcinogenesis in the stomach[27-30]. The best known target of NSAIDs is COX, the rate limiting enzyme in the conversion of arachidonic acid to prostanoids. Thus, the chemo-preventive effects of NSAIDs on gastric cancer may be attributed to inhibition of COX and the resulting decrease in production of PGs[31-33]. However, there are several studies that conflict with this observation. For example, NSAID derivates that lack the ability to inhibit COX (sulindac sulfone) have been shown to inhibit gastrointestinal tumor growth in vivo and in vitro[34-37]. COX-1 and COX-2 null mouse embryofibroblast cells remain sensitive to the anti-proliferative and anti-neoplastic effects of NSAIDs[38]. Furthermore, the concentration of NSAIDs that inhibits the growth of cancer cells in vitro is 10-100 times higher than that required to inhibit COX activity, suggesting the existence of additional cellular targets.

Induction of apoptosis

A substantial body of evidence now supports the idea that the induction of programmed cell death (apoptosis) is one of the main ways by which NSAIDs prevent cancer in a COX-2 dependent manner or a COX-2 independent manner.

COX 2-dependent mechanisms

One of the primary pharmacological properties of the NSAIDs is their ability to inhibit the COX. Analysis of COX expression has shown that COX-2 is increased in gastric cancer, but not in normal mucosa[4,39-42]. These findings support the idea that COX-2 over-expression is important during gastric carcinogenesis. In addition, over-expression of COX-2 can increase production of PGE2, adhesion to the extra-cellular matrix, concentrations of BCL2, and reduce TGF-2 receptor expression and E-cadherin protein. All these changes, which suggest an increased tumorigenic potential, support the notion that COX-2 overexpression alters the biology of cells and plays a part in the transformation process[43,44]. The question of how COX-2 inhibition induces apoptosis has been hotly debated. Several studies have suggested that decreased COX-2 activity can reduce eicosanoids such as the PGs, and in turn affect cell proliferation and apoptosis[45]. So far, there is no definitive evidence that supports the existence of a signaling pathway through which PGs can directly affect apoptosis. However, specific observations provide some clues and warrant mention here. The peroxisome proliferator-activated receptors (PPARs), a family of ligand-activated transcription factors comprising a subset within the steroid receptor family of intracellular receptors, have been shown to modulate the apoptotic response of various cell types[46,47]. It is conceivable that certain COX-2 products may be physiological ligands for the PPARs. Thus, it is hypothesized that PPARs are downstream of COX-2 in gastric carcinogenesis and may play a part in NSAIDs-mediated antineoplastic effects[48-50].

Arachidonic acid may also provide a mechanism underlying COX-2-dependent NSAIDs-induced apoptosis. Arachidonic acid is produced by hydrolysis of phospholipids and triglycerides. The phospholipase A2 cleaves phospholipids and releases arachidonic acid into the cytosol. Once released, it serves as a precursor for PGs production and is a signaling molecule in its own right. Treatment with various NSAIDs results in inhibition of COX-2 and a dramatic increase in the concentration of arachidonic acid, the main substrate for COX-2[45,51]. This buildup of arachidonic acid stimulates the sphingomyelinase to convert sphingomyelin to ceramide, a potent inducer of apoptosis[52-55]. Other studies have shown that the cellular concentration of arachidonic acid is a general mechanism by which apoptosis is regulated[56-59]. It seems that arachidonic acid can alter mitochondrial permeability and cause cytochrome C release, leading to apoptosis[58,60].

COX 2-independent mechanisms

Several recent observations cast doubt on the idea that COX is the sole target of NSAID action. The finding that some NSAIDs can inhibit proliferation and induce cell death in cells that do not express COX, suggests that other targets may play a part in NSAIDs-mediated apoptosis[61].

One potential mechanism involves the transcription factor NF-κB, which promotes cell survival and enhances proliferation. Several investigators have suggested that NSAIDs could promote apoptosis by inhibiting NF-κB signaling pathway[62-64]. The effect that seems to be due to its ability to inhibit the activity of IκB kinase β (IKKβ), an enzyme that activates the NF-κB pathway by phosphorylating the inhibitory subunit of NF-κB (IκB-α) and targeting it for destruction. Subsequently, genes required for cancer cell growth and survival may not be transcribed[65]. Furthermore, Yasui et al[66] reported that sulindac inhibits TNF-α-mediated NF-κB activation and greatly sensitizes human gastric cancer cell line MKN45 to TNF-α. The study data strongly suggest that combination therapy of TNF-alpha with sulindac may sensitize tumor cells to TNF-α and augment its proapoptotic potential. Moreover, Wong et al[67] showed that SC236, a selective COX-2 inhibitor, could inhibit NF-κB mediated gene transcription and binding activity in gastric cancer. This effect occurs through a mechanism independent of COX activity and PG synthesis. Furthermore, unlike aspirin, SC236 affects neither the phosphorylation, degradation, nor the expression of IκB-α, suggesting that the effects of SC236 are independent of IKK activity and IκB-α gene transcription. Instead, SC236 works directly through suppressing the nuclear translocation of RelA/p65. It is possible that SC236 directly targets proteins that facilitate the nuclear translocation of NF-κB[67].

The role of protein kinase C (PKC) and apoptosis-related oncogenes has also attracted more attention. Zhu et al[68] showed that both aspirin and indomethacin induce apoptosis in gastric cancer cells which may be mediated by up-regulation of c-myc proto-oncogene. PKC activation can abrogate the effects of NSAIDs by decreasing c-myc expression. Jiang et al[69] also showed that SC-236 induces apoptosis in AGS cells. SC-236 down-regulates the protein expression and kinase activity of PKC-β(1) in AGS cells, which provides an explanation for COX-independent apoptotic effects of selective COX-2 inhibitors in cultured gastric cancer cells. They also suggested that PKC-β(1) acts as a survival mediator in gastric cancer. Down-regulation of PKC-β(1) by SC-236 may provide new targets for future treatment of gastric cancer[69].

Furthermore, it was reported that the pro-apoptotic proteins Bax and Bak play a role in NSAIDs-induced gastric cancer cell apoptosis[70], suggesting that one of the major pathways mediating the anti-tumor response of aspirin and indomethacin in gastric cancer cells is through the up-regulation of Bax and Bak and activation of caspase-3.

Besides, Wu et al[71] demonstrated that SC-236 induces apoptosis in gastric cancer cells via the up-regulation of 15-lipoxygenase-1 (15-LOX-1), and increases endogenous 13-S-hydroxyoctadecadienoic acid (13-S-HODE), a novel target for chemoprevention effect of NSAIDs.

Regulation of angiogenesis

It is now clear that angiogenesis is critical for the progression of most human cancers[72]. Angiogenesis is a complex process controlled by a balance of angiogenic and angiostatic factors involved in multiple pathways that result in endothelial cell proliferation, differentiation, and organization into a functional network of vascular channels. Among the reported angiogenic factors, vascular endothelial growth factor (VEGF) is the most powerful endothelial cell specific mitogen that plays a key role in the complicated process of angiogenesis.

COX-2 has been shown to directly promote angiogenesis in several different experimental systems[73,74]. Masferrer et al[80] have examined more than 150 samples of different types of human cancers and found that COX-2 is present in the angiogenic vasculature in most of the tumors analyzed. Therefore, COX-2-derived PGs contribute to tumor growth by inducing newly formed blood vessels that sustain tumor cell viability and growth. Furthermore, xenografts of COX-2 over-expressing gastric cancer, when implanted into nude mice, express angiogenic factors in a COX-2-dependent manner, and both angiogenesis and tumor growth can be inhibited in these mice by administration of NSAIDs[23,78].

COX-2 expression has been shown to correlate with microvessel density (MVD), an accepted histologic marker of tumor angiogenesis in several different human malignancies, including gastric cancer[30]. Recent years, the relationships between COX-2 expression, VEGF expression, tumor angiogenesis, and outcome parameters in gastric cancer patients have been reported[75-77]: COX-2 expression significantly correlates with VEGF, and the colocalization of these proteins is most frequent at the advancing edge of cancer cells[75]. MVD is higher in COX-2 and VEGF positive cases than in COX-2 and VEGF negative cases[76]. Co-expression of COX-2 and VEGF, age, lymph node status, and serosal invasion are independent prognostic factors for overall survival in gastric cancer patients[77].

Interestingly, COX-1, the other isozyme expressed in tumor vascular endothelia, participates in tumor angiogenesis, because an anti-sense oligonucleotide of COX-1 suppresses in vitro angiogenesis induced by COX-2-overexpressing cells. Non-specific COX inhibitors reduce the growth and angiogenesis in cancer xenografts by inhibiting COX-1 in vascular endothelial cells, even when the tumor does not express COX-2. These results demonstrate that COX inhibitors suppress angiogenesis and tumor growth by inhibiting expression of angiogenic factors and vascular endothelial cell migration[79].

COX-2 expression and H pylori infection

H pylori infection is also a strong risk factor for the development of gastric cancer via the pathway of atrophy, intestinal metaplasia and dysplasia in chronic infection. Infection with H pylori up-regulates mediators of the inflammatory response, leading to the production of inflammatory cytokines and PGs which themselves may suppress cell-mediated immune responses and promote angiogenesis. These factors may also have effects on cell growth and survival signaling pathways, resulting in induction of cell proliferation and inhibition of apoptosis[81].

H pylori up-regulates COX-2 mRNA expression and PGE₂ synthesis in gastric cancer cells[82], and both pre-malignant and malignant gastric lesions demonstrate high COX-2 expression levels[83]. Therefore, H pylori-induced gastric carcinogenesis may be associated with the elevated expression of COX-2 in gastric epithelium. It is possible that NSAIDs may inhibit the replication and proliferation of H pylori[84,85], neutralize the increased COX-2 expression and PG synthesis and reverse the increased apoptosis and proliferation of epithelial cells associated with

H pylori infection[86,89], thereby reducing the risk of gastric cancer[90,91]. Experimental models of cell proliferation have also shown that, in hyperplastic gastritis in the murine model infected with H pylori, treatment with etodolac could significantly decrease PGE₂ production and the thickness of gastric pits in the infected groups[87].

Whether eradication of H pylori can prevent gastric cancer remains contradictory. COX-2 expression remains in gastric epithelium even after successful eradication of H pylori[88]. This may suggest that combined treatment of COX-2 inhibitors and H pylori infection may be more useful in eliminating H pylori-induced COX-2 expression than either anti-H pylori antibiotics or COX-2 inhibitors alone. This issue deserves further study.

CONCLUSIONS AND FUTURE DIRECTIONS

Combining the evidence from humans, animal models and cell culture systems can establish the targeted inhibition of COX-2 was a viable approach to gastric cancer prevention and/or treatment. However, despite these successes, many questions remain unanswered such as the optimal dose and duration of treatment. Selective COX-2 inhibitors, which have an advantage of minimizing the gastrointestinal side-effects, have received wide attention. An exciting area for future investigation is to investigate whether COX-2 inhibitors can be used in the treatment of patients who have preneoplastic gastric lesions or suffer from gastric cancer, and whether combination therapy with COX-2 inhibitors and the drugs targeting other oncogenic pathways can lead to improved clinical outcomes.