Anti-Helicobacter pylori therapy followed by celecoxib on progression of gastric precancerous lesions

Abstract

AIM: To evaluate whether celecoxib, a selective cyclooxygenase 2 (COX-2) inhibitor, could reduce the severity of gastric precancerous lesions following Helicobacter pylori (H pylori) eradication.

METHODS: H pylori-eradicated patients with gastric precancerous lesions randomly received either celecoxib (n = 30) or placebo (n = 30) for up to 3 mo. COX-2 expression and activity was determined by immunostaining and prostaglandin E₂ (PGE₂) assay, cell proliferation by Ki-67 immunostaining, apoptosis by TUNEL staining and angiogenesis by microvascular density (MVD) assay using CD31 staining.

RESULTS: COX-2 protein expression was significantly increased in gastric precancerous lesions (atrophy, intestinal metaplasia and dysplasia, respectively) compared with chronic gastritis, and was concomitant with an increase in cell proliferation and angiogenesis. A significant improvement in precancerous lesions was observed in patients who received celecoxib compared with those who received placebo (P < 0.001). Of these three changes, 84.6% of sites with dysplasia regressed in patients treated with celecoxib (P = 0.002) compared with 60% in the placebo group, suggesting that celecoxib was effective on the regression of dysplasia. COX-2 protein expression (P < 0.001) and COX-2 activity (P < 0.001) in the gastric tissues were consistently lower in celecoxib-treated patients compared with the placebo-treated subjects. Moreover, it was also shown that celecoxib suppressed cell proliferation (P < 0.01), induced cell apoptosis (P < 0.01) and inhibited angiogenesis with decreased MVD (P < 0.001). However, all of these effects were not seen in placebo-treated subjects. Furthermore, COX-2 inhibition resulted in the up-regulation of PPARγ expression, a protective molecule with anti-neoplastic effects.

CONCLUSION: H pylori eradication therapy followed by celecoxib treatment improves gastric precancerous lesions by inhibiting COX-2 activity, inducing apoptosis, and suppressing cell proliferation and angiogenesis.

Key Words: Apoptosis; Cyclooxygenase 2; Gastric precancerous lesions; Helicobacter pylori; Microvessel density; Proliferation

INTRODUCTION

Gastric cancer is the second leading cause of cancer deaths worldwide[1] and its 5-year survival rate is only 10%-15% in individuals with advanced disease[2]. Helicobacter pylori (H pylori) has been classified as a type I carcinogen by the WHO and is recognized as an important pathogen in gastric tumorigenesis[3]. H pylori infection initiates the inflammatory and atrophic changes in gastric mucosa accompanied by enhanced expression of some protumorigenic factors such as cyclooxygenase 2 (COX-2) and anti-apoptosis proteins, resulting in uncontrolled proliferation of mutated atrophic cells, suppression of apoptosis, excessive angiogenesis and finally the formation of adenocarcinoma[4].

In cancer prevention, the targeting of precancerous lesions has been recognized as the most promising method. However, progress has been achieved only in the chemoprevention of colorectal neoplasia[5]. There is no effective therapy for reversing gastric premalignant lesions. Although H pylori infection is a critical initiator and mediator in gastric premalignant changes and gastric carcinogenesis, eradication of H pylori alone failed to improve these precancerous lesions[6–8]. The failure of H pylori eradication may be explained by the fact that the expression of COX-2, an important mediator in H pylori-induced premalignant changes, remained high or only modestly reduced after H pylori eradication[6–8]. It has been widely accepted that COX-2 plays an important role in gastric carcinogenesis[9]. COX-2 over-expression has been found in H pylori-induced inflammation, precancerous lesions and gastric tumors[9]. Inhibition of COX-2 by non-steroidal anti-inflammatory drugs (NSAIDs) has been proven to be effective in preventing gastric carcinogenesis as evidenced in animal models and epidemiological studies[2910]. Since eradication of H pylori alone is not sufficient to reverse gastric carcinogenesis due to failure of the inhibition of H pylori-induced protumorigenic factors such as COX-2, it would be reasonable to combine additional treatments such as COX-2 inhibition following H pylori eradication.

In the present study, we examined whether treatment with the selective COX-2 inhibitor celecoxib could reduce the severity of gastric precancerous lesions after H pylori eradication. The mechanisms of its action were also investigated.

MATERIALS AND METHODS

Patients and study design

We enrolled 233 patients who had upper-gastrointestinal symptoms such as anorexia, early satiety, stomach pain, abdominal distention and epigastric discomfort between January 2005 and July 2006 from the First Affiliated Hospital of Hebei Medical University, Shijiazhuang, China. Eligible subjects were between 30 and 70 years of age, and had no history of drug allergies. Subjects were ineligible if they were under 30 or older than 70 years old; pregnant or lactating; had peptic ulcer; gastric cancer or other cancers; upper gastrointestinal bleeding; liver cirrhosis; serious cardiovascular diseases, renal or lung diseases; hypersensitive to COX-2 inhibitors, NSAIDs, salicylates, or sulfonamides; used NSAIDs; and those unwilling to undergo repeat endoscopy after treatment.

H pylori infected patients who were defined both by the presence of the bacterium on histology as well as a positive C¹⁴ urea breath test were treated with a 1-wk course of eradication therapy (omeprazole, 20 mg; amoxicillin, 1000 mg; furaltadone, 100 mg; twice daily). Five weeks post-treatment, patients were recalled to hospital for repeat endoscopy and C¹⁴ urea breath test. H pylori eradication was confirmed by a negative C¹⁴ urea breath test and a negative H pylori histology examination from biopsies. Only patients with confirmed H pylori eradication were recruited into this study. During each endoscopy, a total of eight gastric biopsies were taken from the antrum (two from the greater curvature and two from the lesser curvature) and the corpus of the stomach (two from the greater curvature and two from the lesser curvature) for molecular experiments and histological examination. One of the two biopsy specimens from each site was immediately stored at -80°C for RNA/protein extraction. The other specimens were fixed in 10% neutral buffered formalin and embedded in paraffin for histological examination and immunostaining. The severity of gastric histology including gastritis and precancerous lesions [gastric atrophy, intestinal metaplasia (IM) and low-grade dysplasia] was scored based on the following standards: absent (0), mild (1), moderate (2) and marked (3)[11]. Patients with high-grade dysplasia were excluded from the study. All gastric biopsies were interpreted by two pathologists who were unaware of the treatment assignments.

One hundred and thirty six eligible subjects who were histologically confirmed to have gastric precancerous lesions and negative tests for H pylori were randomly assigned to receive either celecoxib 200 mg twice daily or an identical-looking placebo at a 1:1 ratio for 3 mo. During the treatment period, 10 subjects were withdrawn due to adverse events and 66 were lost to follow-up. Some patients lost to follow-up migrated to other cities, some were non-compliant and failed to follow through with our treatment protocol, and some refused to undergo a second endoscopy. At the end of the 3-mo treatment period, 30 patients in the celecoxib group (16 males with an average age of 50.33 ± 10.39 years; 14 females with an average age of 57.36 ± 9.25 years) and 30 patients in the placebo group (14 males with an average age of 48 ± 10.43 years; 16 females with an average age of 50 ± 9.98 years) returned for endoscopic examination. The same protocols for obtaining gastric biopsy and histological examination were used as previously mentioned in the baseline endoscopy. The study protocol was approved by the Clinical Research Ethics Committee and all participants gave written informed consent.

Immunohistochemistry

The paraffin-embedded gastric sections were deparaffinized and rehydrated in graded ethanol. The activity of endogenous peroxidase was blocked by methanol containing 3% H₂O₂ for 10 min and washed with PBS. After blocking with 10% nonimmunized goat serum at 37°C for 20 min, sections were incubated with the primary antibody for COX-2 (1:500, Santa Cruz Biotechnology, Santa Cruz, CA, USA), Ki-67 (1:200, Chemicon International Inc., Temecula, CA, USA), peroxisome proliferator-activated receptor γ (PPARγ) (1:100, Santa Cruz) overnight at 4°C. Peroxidase activity was visualized by applying diaminobenzidine to the sections, which were then counter-stained with hematoxylin. Analysis of the immunostained sections was independently performed by two pathologists in a blinded fashion.

Microvessel density (MVD) was performed on paraffin-embedded gastric tissue sections stained with CD31 (1:150, DAKO, Glostrup, Denmark) as an indicator. For the determination of MVD in each case, five of the most highly vascularized areas within a section were selected and counted under a light microscope. The average numbers of microvessels in the selected fields were recorded as the MVD for each case.

cDNA synthesis and RT-PCR

Gastric tissue specimens were homogenized with a homogenizer. Total RNA was extracted using TRIzol reagent (Invitrogen, USA). Five micrograms of total RNA from each sample was reverse transcribed into cDNA using the AMV Reverse Transcription system (Promega, San Luis Obispo, CA, USA). Semi-quantitative PCR was performed. The primer sequences of proliferation cell nuclear antigen (PCNA), factor associated suicide (FAS), PPARγ and β-actin are listed in Table 1. A DNA free template control (containing water) was included and each sample was added in duplicate. PCR products were separated by 15% agarose gel electrophoresis and quantified by the Gel Imaging System after ethidium bromide (10 mg/L) staining. The mRNA expression level of the target gene was defined by the densitometry ratio of target gene to β-actin.

Table 1 Primer sequences used for amplification of mRNA by semi-quantitative PCR.

mRNA	Forward primer (5’-3’)	Reverse primer (5’-3’)	Size (bp)
PCNA	CTTTTCTGTCACCAAATTTGTACC	AACTGCATTTAGAGTCAAGACCC	206
Fas	CTGCCAAGAAGGGAAGGAGT	GGTGCAAGGGTCACAGTGTT	189
PPARγ	AGCCTCATGAAGAGCCTTCCA	ACCCTTGCATCCTTCACAAGC	89
β-actin	GTCTTCCCCTCCATCGTG	GGGTGAGGATGCCTCTCTT	251

PGE₂ assay

PGE₂ levels were measured in snap frozen tissue specimens using a radioimmunoassay-based assay. Briefly, about 20 mg of snap frozen tissues were homogenized in 10 volumes of sodium chloride by a ground glass homogenizer on ice. The mixture was incubated at 37°C for 15 min and then centrifuged for 20 min at 3000 r/min. The supernatant was then applied to the pre-primed immunoassay reaction mixture and reacted with the antibody overnight. PGE₂ was precipitated with 0.7 mL volumes of 25% polyethylene glycol. The quantity of PGE₂ in the supernatants was determined using RIA.

Determination of cell proliferation

Proliferation was assayed by immunoperoxidase staining for Ki-67 as described previously[12]. The immunochemistry method for staining the sections with Ki-67 antibody has been specified above. The proliferation index (PI) was defined as a percentage of the ratio of Ki-67-positive nuclei to the total nuclei counted.

Terminal deoxynucleotidyl transferase-mediated nick end labeling (TUNEL)

Apoptosis was determined in situ from paraffin-embedded tissue sections by the TUNEL assay using the in situ Cell Death Detection kit (Roche Applied Science, Indianapolis, IN, USA). Briefly, paraffin-embedded slides were deparaffinized, hydrated and incubated with proteinase K (20 mg/mL in 10 mmol/L Tris-HCl) for 20 min. After labeling with the TUNEL reaction mixture, slides were developed by converter-POD and DAB substrate. PBS replaced the primary antibodies as a negative control. The results of staining were analyzed and evaluated by three individuals independently. At least 1000 cells were counted in five random fields. Apoptosis index (AI) was represented as the percentage of positive cells with TUNEL staining to the total cells.

Statistical analysis

Data were analyzed by SPSS 12.0 software and are shown in a default form of mean ± SD. The association between COX-2 expression and the progression of precancerous lesions was analyzed by χ² test. The correlation between COX-2 expression and MVD/PI was analyzed using Pearson’s Correlation Coefficient. The difference between the groups was compared by t test. The paired t test was used to determine the difference between pre-treatment and after-treatment within each treatment group. P < 0.05 was considered statistically significant.

RESULTS

Association between COX-2 expression and progression of precancerous lesions

The percentage of COX-2 positive cells in gastric tissues was increased with the progression of chronic gastritis (33.3%), atrophy (51.6%), IM (53.3%), and dysplasia (79.3%) as determined by immunostaining. COX-2 expression in dysplasia (79.3%) was significantly higher than in any other type of lesion (P < 0.05). COX-2 expression was significantly elevated in atrophy and IM compared with chronic gastritis (P < 0.05). However, there was no significant difference between gastric atrophy and IM (P > 0.05).

COX-2 expression correlated with cell proliferation

Cell proliferation was determined by Ki-67 staining. A significant positive correlation between COX-2 expression and PI (χ² = 10.5, P = 0.001) was observed by Pearson’s Correlation Coefficient analysis, indicating COX-2 expression correlated with cell proliferation.

COX-2 expression correlated with angiogenesis

We also evaluated the correlation between COX-2 expression and angiogenesis as determined by CD31 immunostaining. The mean MVD was significantly higher in COX-2-positive tissues (n = 47, 23.85 ± 7.44) than in COX-2-negative tissues (n = 27, 18.47 ± 6.02) (P < 0.001), indicating a positive correlation between COX-2 expression and MVD. Thus COX-2 also played an important role in angiogenesis.

Effect of celecoxib on the improvement in histology of gastric precancerous lesions

After three months treatment, a significant improvement in precancerous lesions was observed in 66.7% (20/30) of patients (P < 0.001) who were treated with celecoxib. However, only 16.1% of cases who received placebo showed improved histology (P > 0.05). Of these three changes, 84.6% of sites with dysplasia regressed in patients treated with celecoxib (P = 0.002) compared with 60% in the placebo group, suggesting that celecoxib was effective on the regression of dysplasia (Table 2). However, differences in pathological improvement of atrophy and intestinal metaplasia were not observed between the celecoxib and placebo groups (Table 2). With regard to the mixed pathological sites with both atrophy and intestinal metaplasia, we did not have sufficient sample size to make an accurate conclusion.

Table 2 Effect of celecoxib on the histological improvement of gastric precancerous lesions.

Precancerous lesions	Regression site n (%)		P-value
	Celecoxib group	Placebo group
Atrophy	11/13 (84.6)	7/13 (53.8)	0.202
Metaplasia	9/14 (64.3)	9/18 (50)	0.419
Atrophy + metaplasia1	3/4 (75)	3/6 (50)	0.895
Dysplasia	35/39 (89.7)	24/40 (60)	0.002

¹Indicates that both atrophy and metaplasia were regressed in one site.

Effect of celecoxib on COX-2 protein expression and COX-2 activity

COX-2 protein expression as determined by the percentage of COX-2 positive cells in the gastric tissues was significantly lower after treatment with celecoxib (pre-treatment: 40.93% ± 11.96% vs after-treatment: 27.88% ± 4.94%, P < 0.001) (Figure 1A-C). COX-2 protein expression was reduced in 73%, remained the same in 13% and increased in 13% of patients treated with celecoxib. However, COX-2 expression was reduced in only 48% of patients in the placebo group (Table 3). PGE₂ level, an indicator of COX-2 activity, was concomitantly reduced in patients treated with celecoxib (pre-treatment: 358.9 ± 59.3 vs after-treatment: 143.6 ± 24.2, P < 0.001). However, these differences in COX-2 expression and COX-2 activity were not seen in patients treated with placebo (Figure 1D).

Figure 1 Effects of celecoxib on COX-2 expression and PGE₂ levels. A: Representative image of COX-2 protein expression as determined by immunostaining in paraffin-embedded gastric tissue sections: (A1) pre-treatment, and (A2) post-treatment with celecoxib; B: Percentage of COX-2 positive cells in gastric mucosa; C: PGE₂ levels. Data are mean ± SD. ^bP < 0.001.

Table 3 Effect of celecoxib on COX-2 expression and other parameters.

	Celecoxib group n (%)			Placebo group n (%)
	Reduction	Same	Worse	Reduction	Same	Worse
COX-2 expression	22 (73)	4 (13)	4 (13)	15 (48)	3 (10)	13 (43)
Cell proliferation	23 (77)	2 (7)	5 (17)	11 (37)	3 (10)	16 (53)
Cell apoptosis	8 (53)	4 (27)	3 (20)	4 (27)	5 (17)	6 (40)
Angiogenesis	22 (73)	4 (13)	4 (13)	12 (39)	10 (32)	9 (29)
PPARγ expression	15 (50)	5 (17)	10 (33)	9 (29)	5 (16)	17 (55)

Effect of celecoxib on cell proliferation

Cell proliferation was significantly reduced in the celecoxib group after 3 mo of treatment (pre-treatment: 27.46% ± 6.77% vs after-treatment: 20.18% ± 4.05%, P < 0.01) (Figure 2A and B). However, this difference was not observed in patients treated with placebo (Figure 2B). Seventy percent of patients in the celecoxib group had reduced PCNA protein expression while a reduction in PCNA expression was observed in only 37% of patients in the placebo group (Table 3). In addition, mRNA expression of PCNA was markedly down-regulated by celecoxib treatment (Figure 2C).

Figure 2 Effects of celecoxib on cell proliferation. A: Representative Ki-67 staining of gastric mucosa before (A1) and after treatment (A2) with celecoxib; B: Celecoxib led to a significant reduction in proliferation index (PI); C: Celecoxib down-regulated mRNA expression of proliferation cell nuclear antigen (PCNA). 1 and 2: Plecebo; 3 and 4: Celecoxib. Data are mean ± SD. ^bP < 0.01.

Effect of celecoxib on cell apoptosis

Treatment with celecoxib significantly induced cell apoptosis as assayed by TUNEL staining (pre-treatment: 3.86% ± 0.44% vs after-treatment: 7.72% ± 0.64%, P < 0.01) (Figure 3A, Table 3). In keeping with this, mRNA expression of the pro-apoptotic gene fas was also enhanced (Figure 3B), suggesting that celecoxib-induced apoptosis was via up-regulation of fas expression.

Figure 3 Effects of celecoxib on cell apoptosis. A: Celecoxib induced the apoptosis index (AI) in gastric mucosa; B: Celecoxib up-regulated mRNA expression of Fas, a pro-apoptosis gene. 1 and 2: Plecebo; 3 and 4: Celecoxib. Data are mean ± SD. ^bP < 0.01.

Effect of celecoxib on angiogenesis

Angiogenesis was evaluated by the MVD assay using CD31 staining. MVD was significantly lower after celecoxib treatment (P < 0.001) (Figure 4, Table 3), indicating a suppressive effect on angiogenesis by celecoxib. However, this difference was not seen in the placebo group.

Figure 4 Effects of celecoxib on angiogenesis. A: Representative microvessel image of paraffin-embedded gastric tissue sections stained with CD31. (A1) pre-treatment, and (A2) post-treatment with celecoxib; B: Celecoxib suppressed microvessel density (MVD). Data are mean ± SD. ^bP < 0.001.

Effect of celecoxib on PPARγ expression

Celecoxib treatment led to an increase in the number of PPARγ positive cells as determined by immunostaining (pre-treatment: 18% ± 4.33% vs after-treatment: 22.6% ± 4.3%, P < 0.05) (Figure 5, Table 3). Thus, celecoxib resulted in up-regulation of PPARγ expression.

Figure 5 Effects of celecoxib on PPARγ protein expression. Representative PPARγ expression in paraffin-embedded gastric tissue sections by immunostaining. (A1) pre-treatment, and (A2) post-treatment with celecoxib. Nuclear staining of PPARγ was markedly increased post-treatment with celecoxib. ^aP < 0.05.

DISCUSSION

We report here that COX-2 is markedly up-regulated in gastric tissues with inflammation and was more prominent during progression in precancerous lesions in H pylori-eradicated patients. Moreover, the induction of COX-2 appeared to coincide with increased cell proliferation and angiogenesis. These results suggested that COX-2 was induced by H pylori infection and mediated by H pylori-associated premalignant gastric lesions. Our observation on the profile of COX-2 expression is supported by previous studies[7813–16], suggesting that COX-2 is a relatively early event and plays an important role during gastric carcinogenesis.

To evaluate whether progression of precancerous lesions could be reduced or reversed, we conducted a 3-mo intervention of celecoxib in patients with precancerous lesions after H pylori eradication. Our results showed that the histology of precancerous lesions was improved in 66.7% of patients treated with celecoxib, which was significantly higher than the placebo group (16.1%) (P < 0.001). Of the three precancerous changes (atrophy, intestinal metaplasia and dysplasia), celecoxib was effective on the regression of dysplasia. However, the evidence for chemopreventive effects on gastric precancerous lesions by NSAIDs has only been limited to animal experiments[1317–22]. In the animal model of carcinogenesis induced by co-treatment with H pylori and N-methyl-N-nitrosourea (MNU), mice underwent H pylori-induced gastritis with multifocal atrophy and intestinal metaplasia, and finally gastric adenocarcinoma. Long term co-administration with a COX-2 inhibitor, either celecoxib or nimesulide not only reduced the development of intestinal metaplasia, but also adenocarcinoma[20–22]. In human studies, a recent report showed that treatment with NSAIDs for more than 3 mo reversed H pylori-induced harmful effects in gastric epithelial cells[23]. On the other hand, large-scale clinical trials have shown that NSAIDs were effective in the chemoprevention of colorectal neoplasia[5]. NSAIDs (celecoxib and sulindac) promoted regression in both number and size in high risk individuals with familial adenomatous polyposis[5]. In the more common sporadic setting, refecoxib and celecoxib reduced the occurrence of human colorectal adenomas[5]. Collectively, these results suggest that the protective effects of NSAIDs such as celecoxib could effectively prevent or reverse the precancerous lesions of gastric cancer.

We further evaluated the underlying mechanisms for the anti-tumorigenic effects of the COX-2 inhibitor. On the basis of the cell proliferation and apoptosis analyses, improvements in precancerous lesions caused by celecoxib were most likely associated with the suppression of cell proliferation and induction of cell apoptosis. In keeping with this, in an animal experiment, celecoxib but not indomethacin suppressed gastric cancer formation by inducing cell apoptosis and suppressing cell proliferation[2425] in a dose dependent manner. A recent report also indicated that NSAIDs induced apoptosis through activation of extrinsic and intrinsic pathways of apoptosis[26]. However, the pro-apoptotic efficacy of various NSAIDs differed greatly[2728]. That may explain why the anti-tumor effects of NSAIDs varied in different experiments.

In the present work, we showed that, in addition to inhibition of cell proliferation and induction of apoptosis, the regression of precancerous lesions in the stomach by celecoxib was also related to inhibition of angiogenesis. It is well-established that the inducible enzyme COX-2 is an important mediator of angiogenesis during tumor growth[29]. COX-2 expression significantly correlated with MVD[30] and vascular endothelial growth factor (VEGF) in human gastric adenomas and carcinomas[31]. The pro-angiogenic effects of COX-2 are mediated primarily by the metabolites of arachidonic acid, resulting in increased production of VEGF, enhanced survival of endothelial cells, induction of matrix metalloproteinases, promotion of vascular sprouting and migration and activation of epidermal growth factor receptor-mediated angiogenesis[29]. In this regard, we showed in this study that induction of COX-2 is parallel with the induced angiogenesis in precancerous lesions. Treatment with celecoxib inhibited angiogenesis with a concomitant histological regression of precancerous lesions. Others have also reported that NSAIDs including celecoxib inhibited angiogenesis and decreased tumor growth in gastric cancer and other cancers both in vitro and in vivo in animal models[1332–34]. Our study provided the first clinical evidence that treatment with celecoxib effectively suppressed angiogenesis and lowered MVD in H pylori-eradicated patients with gastric precancerous lesions.

The expression of PPARγ, a protective anti-neoplastic molecule[35], was enhanced by the COX-2 inhibitor rofecoxib in human gastric cancer with a concomitant induction of apoptosis and attenuation of proinflammatory cytokines production[36]. We also observed the up-regulation of PPARγ in patients treated with celecoxib. The up-regulation of PPARγ by NSAIDs was reported either via the COX-2 independent pathway or the COX-2 dependent pathway[37–39]. Activation of PPARγ was shown by us and others to prevent mammary carcinogenesis in experimental animals[40–42] through suppression of COX-2 expression[40]. In mice treated with MNU and H pylori, nimesulide administration substantially reduced H pylori-associated gastric tumorigenesis along with substantial activation of PPARγ and induction of apoptosis[21]. Collectively, these findings raise the possibility that up-regulation of PPARγ by celecoxib contributed to the histological improvement in precancerous lesions.

In conclusion, COX-2 expression was induced in gastric epithelium with the progression of precancerous lesions. Eradication of H pylori combined with a 3-mo intervention of celecoxib was effective in improving the severity of precancerous lesions mainly by inducing apoptosis, and inhibiting cell proliferation and angiogenesis. Thus COX-2 is a promising target in reversing gastric precancerous lesions and celecoxib showed efficacy in the chemoprevention of these lesions.

COMMENTS

Background

Epidemiologic studies have shown that cyclooxygenase 2 (COX-2) inhibitor could reduce the risk of gastric cancer. The authors aim to evaluate whether celecoxib, a selective COX-2 inhibitor, could reduce the severity of gastric precancerous lesions following Helicobacter pylori (H pylori) eradication.

Research frontiers

Gastric cancer is the most common cancer and the leading cause of cancer-related death in China, with an overall 5-year survival rate of only 10%-20%. There is a compelling need to explore the novel targets that contribute to gastric carcinogenesis for effective treatment. The development of gastric cancer is generally believed to be a multi-step progression from chronic gastritis to atrophy, intestinal metaplasia (IM), dysplasia and cancer, that is triggered by H pylori infection. Eradication of H pylori alone is not efficient in preventing the progression of gastric IM. The authors hypothesized in addition to eradicate H pylori, inhibition of COX-2, a potential oncogene gene that was induced in the early stage of gastric carcinogenesis, by selective COX-2 inhibitor (celecoxib) may regress the premalignant changes in the stomach by suppressing COX-2. Herein, they tested the effect of a specific COX-2 inhibitor in patients with confirmed gastric atrophy and/or IM after H pylori eradication. The present study was a prospective, randomized, and placebo-controlled study.

Innovations and breakthroughs

The authors have demonstrated in this study that H pylori eradication therapy followed by celecoxib treatment improves and dampens the progression of gastric precancerous lesions. The anti-neoplastic properties of celecoxib were due to its ability of inhibiting COX-2 activity, inducing apoptosis, suppressing cell proliferation and angiogenesis.

Applications

COX-2 was a promising target in reversing gastric precancerous lesions and celecoxib showed efficacy in this chemoprevention. This finding may provide clinical implication.

Peer review

The authors report interesting data on the effect of celecoxib administration in patients with gastric precancerous lesions following H pylori eradicated. This is a well designed study with interesting results.