© The Author 2006. Published by Oxford University Press.
ARTICLE |
Randomized Double-Blind Factorial Trial of Three Treatments To Reduce the Prevalence of Precancerous Gastric Lesions
Affiliations of authors: Peking University School of Oncology, Beijing Cancer Hospital and Beijing Institute for Cancer Research, Beijing, China (WY, LZ, JL, MJ, YC, JM, KP, GX); Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, MD (LMB, JFF, MHG); Linqu County Public Health Bureau, Shandong, China (WL); Westat, Rockville, MD (YH, SCM); Information Management Services, Rockville, MD (DP); International Epidemiology Institute, Ltd., Rockville, MD, and Vanderbilt University, Nashville, TN (WJB)
Correspondence to: Mitchell Gail, MD, PhD, National Cancer Institute, 6120 Executive Blvd., EPS 8032, Bethesda, MD 20892-7244 (e-mail: gailm{at}mail.nih.gov).
 |
ABSTRACT |
|---|
 TopNotes Abstract IntroductionPopulation and methodsResultsDiscussionReferences |
|---|
Background: Randomized trials have yielded mixed results on the effects of treatment for Helicobacter pylori and little information on the effects of vitamins or garlic supplements on precancerous gastric lesions. We conducted a randomized trial to test the effects of one-time H. pylori treatment and long-term vitamin or garlic supplements in reducing the prevalence of advanced precancerous gastric lesions. Methods: Most of the adults aged 35–64 years in 13 randomly selected villages in Linqu County, Shandong Province, China, were identified and given baseline endoscopies in 1994. In 1995, 3365 eligible subjects were randomly assigned in a factorial design to three interventions or placebos: amoxicillin and omeprazole for 2 weeks in 1995 (H. pylori treatment); vitamin C, vitamin E, and selenium for 7.3 years (vitamin supplement); and aged garlic extract and steam-distilled garlic oil for 7.3 years (garlic supplement). Subjects underwent endoscopies with biopsies in 1999 and 2003, and the prevalence of precancerous gastric lesions was determined by histopathologic examination of seven standard biopsy sites. The 3365 eligible randomized subjects represented 93.5% of those with baseline endoscopy and included all baseline histologic categories except gastric cancer. Only 0.18% had normal gastric mucosa. Logistic regression was used to estimate the intervention effects on the odds of advanced precancerous gastric lesions, and t-tests were used to assess effects on histologic severity. All statistical tests were two-sided. Results: H. pylori treatment resulted in statistically significant decreases in the combined prevalence of severe chronic atrophic gastritis, intestinal metaplasia, dysplasia, or gastric cancer in 1999 (odds ratio [OR] = 0.77; 95% confidence interval [CI] = 0.62 to 0.95) and in 2003 (OR = 0.60; 95% CI = 0.47 to 0.75), and had favorable effects on the average histopathologic severity and on progression and regression of precancerous gastric lesions in 2003. H. pylori treatment did not reduce the combined prevalence of dysplasia or gastric cancer. However, fewer subjects receiving H. pylori treatment (19/1130; 1.7%) than receiving placebo (27/1128; 2.4%) developed gastric cancer (adjusted P = .14). No statistically significant favorable effects were seen for garlic or vitamin supplements. Conclusion: H. pylori treatment reduces the prevalence of precancerous gastric lesions and may reduce gastric cancer incidence, but further data are needed to prove the latter point. Long-term vitamin or garlic supplementation had no beneficial effects on the prevalence of precancerous gastric lesions or on gastric cancer incidence.
|
INTRODUCTION |
|---|
|
TopNotesAbstractIntroductionPopulation and methodsResultsDiscussionReferences |
|---|
Gastric cancer is the second leading cause of cancer mortality worldwide (1) and accounts for 42% of the cancer deaths in Linqu County, Shandong Province, China (2). The prevalence of precancerous gastric lesions is very high in Linqu County (3), where 67% of adults have antibodies to Helicobacter pylori (4). This bacterium is found in the gastric mucous layer or adherent to the epithelial lining of the stomach and is thought to cause gastric cancer (5). A longitudinal study in Linqu County (6) found that the relative risks for subsequent development of gastric cancer were 104 (95% confidence interval [CI] = 9.7 to 999) for subjects with moderate or severe dysplasia, 29 (95% CI = 3.9 to 219) for subjects with mild dysplasia or deep intestinal metaplasia, and 17 (95% CI = 1.5 to 202) for subjects with superficial intestinal metaplasia, all compared with subjects with superficial gastritis or chronic atrophic gastritis. These results, along with histopathologic evidence for the multistep nature of gastric carcinogenesis (7), suggest that interventions that slow the progression of precancerous gastric lesions may reduce the incidence of gastric cancer.
We conducted a randomized, placebo-controlled factorial-design trial of one-time antibiotic treatment for H. pylori infection and/or 7.3 years oral supplementation with a vitamin preparation and/or 7.3 years oral supplementation with a garlic preparation to evaluate effects on the prevalence of advanced precancerous gastric lesions. We used the combination of amoxicillin and omeprazole to treat H. pylori infection because it had been proven safe and effective in a pilot study (4); this treatment was administered to H. pylori–seropositive subjects only. We also evaluated the effect of a supplement consisting of vitamins E and C and selenium because observational data from Linqu County indicated that increasing consumption of fruits and vegetables was associated with a decreasing risk of gastric cancer (8) and that a higher serum level of vitamin C was associated with less advanced precancerous gastric lesions (9) and less chance of progression of precancerous gastric lesions (10). In addition, an earlier trial in Henan Province, which also has a high incidence of gastric cancer, indicated that a mixture of 
-carotene, vitamin E, and selenium reduced gastric cancer mortality (11). Finally, we also evaluated the effect of a mixture of aqueous ethanol–aged garlic extract and steam-distilled garlic oil because increased consumption of allium-containing vegetables, especially garlic, has been associated with a decreased risk of gastric cancer in Linqu County (12).
|
POPULATION AND METHODS |
|---|
|
TopNotesAbstractIntroductionPopulation and methodsResultsDiscussionReferences |
|---|
Study Population
Details of the study design and population have been described previously (4). Briefly, in 1994, a census of 13 randomly selected villages within four townships in Linqu County, Shandong Province, China, identified 4010 residents who were aged 35–64 years. In 1994, 3599 of these residents agreed to undergo a gastroscopy with biopsy and to provide blood for serology to detect H. pylori infection at baseline. In the summer of 1995, these individuals were invited to participate in the current trial and provided written informed consent. A total of 3411 subjects who were thought to be eligible were randomly assigned to treatment groups in a stratified factorial design (Fig. 1 and Table 1). Eligibility exclusions before randomization included refusal to provide informed consent, residence outside the study villages, penicillin allergy, missing baseline H. pylori serology, ages outside the range of 35–64 years, previous H. pylori treatment, previous diagnosis of cancer (except nonmelanoma skin cancer), bleeding disorder, heart failure, emphysema, renal or liver disease, or other life-threatening illness. After randomization, it was determined that 46 subjects were actually ineligible before randomization (two subjects had died, 38 subjects had been diagnosed with gastric cancer, and six subjects had had been diagnosed with another cancer), leaving 3365 subjects available for analysis (Fig. 1 and Table 1). These determinations were made independently of treatment received or outcome; although ineligible, 41 of these 46 subjects received interventions. Follow-up endoscopies with biopsies were performed between March 15, 1999, and May 3, 1999, and between March 16, 2003, and April 24, 2003. This study was approved by the institutional review boards of the Beijing Institute for Cancer Research, the U.S. National Cancer Institute, and Westat, and written informed consent to participate was obtained from each subject. This trial is registered in the U.S. National Cancer Institute PDQ database (trial number NCI-OH-95-C-N029; available at http://www.cancer.gov/clinicaltrials/).
|
|
Study Design and Randomization
Subjects were stratified on the basis of their H. pylori antibody status at baseline in 1994 (H. pylori–seropositive versus H. pylori–seronegative) (4). H. pylori–seropositive subjects were randomly assigned to three interventions (antibiotics and/or garlic supplements and/or vitamin supplements) or their placebos in a 2 x 2 x 2 factorial design (Fig. 1 and Table 1). H. pylori–seronegative subjects were randomly assigned to vitamin supplements and/or garlic supplements in a 2 x 2 factorial design; these subjects also received placebo for the initial H. pylori treatment to protect blinding (Fig. 1 and Table 1). To ensure treatment balance with respect to sex and age, subjects were ordered by age within four groups defined by sex and by H. pylori serostatus in 1994. Among H. pylori–seropositive males and H. pylori–seropositive females, the ordered subjects were assigned at random, in blocks of eight, to the eight treatment combinations (Table 1), and among H. pylori–seronegative males and H. pylori–seronegative females, the subjects were assigned at random, in blocks of four, to the four treatment combinations (Table 1). Both the participants and the investigators were masked to treatment assignment (4). Randomized treatment assignments were generated at Westat in the United States after eligibility was determined. Pill bottles bearing codes corresponding to those assignments were then distributed to the study participants in Linqu County.
Treatments
From September 15 to November 29, 1995, eligible H. pylori–seropositive subjects were given capsules containing amoxicillin (1 g) and omeprazole (20 mg) (N = 1130) or placebo (N = 1128) to take twice daily for 2 weeks. The 2-week course of active treatment was offered again to the 382 subjects for whom the initial course of active therapy did not eradicate H. pylori, as determined by results of [13C]urea breath tests (CUBT) (which detect the activity of urease, an enzyme produced by H. pylori) that were conducted from January to March 1996. To preserve masking, 383 subjects in the placebo group who were matched on village, age, and sex were offered re-treatment with placebo. Vitamins and garlic supplementation began on November 30, 1995, and continued until March 31, 2003, for a total of 7.3 years. The vitamin supplement (N = 1677) was a capsule that contained vitamin C (250 mg), vitamin E (100 IU), and selenium from yeast (37.5 µg) or placebo (N = 1688); it was to be taken twice daily. (We refer to this combination as "vitamins," even though it included selenium.) The garlic supplement (N = 1678) was a capsule that contained 200 mg of aged garlic extract (known as Kyolic and produced by Wakunaga Pharmaceutical Co., Osaka, Japan) and steam-distilled garlic oil (1 mg) or placebo (N = 1687); two such capsules were to be taken twice daily. Bottles containing the placebo capsules for the garlic supplement contained trace amounts of steam-distilled garlic oil to preserve masking. The placebos are described in detail elsewhere (4).
Gastroscopy and Histopathology
Gastroscopy procedures, including biopsy samples taken at seven standard sites in the stomach, and histopathologic criteria have been described elsewhere (3,6). The gastroenterologists and pathologists were blinded to the subjects' interventions. Quality-control studies were conducted during the trial by Dr. J.-y. Li and by two outside advisors, Drs. P. Correa of Vanderbilt University and P. Sipponen of Helsinki University Central Hospital, to assess agreement among readers and comparability of the Chinese histopathology grading criteria with those of the updated Sydney System (USS) (13). The Chinese criteria, which were used in this study, corresponded to the USS grading system, with the following exceptions: superficial intestinal metaplasia in the Chinese system corresponded to mild intestinal metaplasia in USS, deep intestinal metaplasia to moderate or marked intestinal metaplasia in USS, mild dysplasia to low-grade dysplasia in USS, and severe dysplasia to high-grade dysplasia in USS. Before the trial began, Drs. Li and Correa analyzed 270 baseline biopsy samples. They assigned the same histopathologic classification in 267 cases and differed only as to whether dysplasia was mild or borderline in three cases (6).
Endpoints and Statistical Analysis
At baseline (in 1994) and in 1999 and 2003, each biopsy site was assigned a severity score according to its histopathologic diagnosis in the Chinese system: 0 for normal, 1 for superficial gastritis, 2 for mild/moderate chronic atrophic gastritis, 3 for severe chronic atrophic gastritis, 4 for superficial intestinal metaplasia), 5 for deep intestinal metaplasia, 6 for mild dysplasia, 7 for moderate dysplasia, 8 for severe dysplasia, and 9 for gastric cancer. Each subject was assigned a global severity score in 1999, defined as the highest severity score in any biopsy site in 1999; a global severity score was again assigned in 2003. The global severity score was upgraded to 9 (gastric cancer) if a subject had a diagnosis of gastric cancer that was documented in a cancer abstract that was based on clinical or pathologic data obtained other than at the scheduled endoscopies in 1999 and 2003. A cancer abstract form was filed to document all incident cancers and included diagnostic code, basis of diagnosis, date and place of diagnosis, and location in the stomach if the cancer was a stomach cancer. Hereafter, we refer to the global severity score simply as the severity score. The three endpoints specified in the study protocol for analysis in 1999 and 2003 were 1) prevalence of dysplasia or gastric cancer (i.e., severity score 
6); 2) prevalence of severe chronic atrophic gastritis, intestinal metaplasia, dysplasia, or gastric cancer (i.e., severity score 3); and 3) average severity score.
This study was designed to achieve a statistical power of 0.96 (for the garlic and vitamin interventions) and of 0.88 (for the antibiotic intervention) to detect a 5% decrease in the prevalence of dysplasia or gastric cancer (i.e., the prevalence of a severity score 6). Although this trial was not designed to detect an effect on gastric cancer incidence, it nonetheless yielded more information on gastric cancer incidence than all previously published trials of H. pylori treatments. We therefore also analyzed gastric cancer incidence. Except for one case, gastric cancer incidence was diagnosed from findings based on endoscopies with biopsies in 1999 and 2003 and from other pre-mortem clinical or pathologic data documented on a cancer abstract. We counted one person with a death abstract indicating gastric cancer but no pre-mortem data on a cancer abstract.
Conditional logistic regression analysis was used to adjust analyses for endpoints 1 and 2 (prevalence of severity score 6 and of severity score 3, respectively). Our analyses focused on the main effects of the interventions, as is customary for studies with factorial designs. However, we also examined interactions among the interventions. For conditional logistic regression analyses of subjects who were H. pylori–seropositive in 1994 (Table 1), we used the following main effects, coded as indicator variables: 1994 histologic severity score (0–2, 3–4, 5, and 6–8), age in 1994 (35–39, 40–44, 45–54, and 55), sex, and the three interventions. The categories for severity score (see Table 3) and age were chosen to yield adequate numbers in each group. Subjects who were H. pylori–seronegative in 1994 were analyzed in a similar fashion except that the indicator for the amoxicillin/omeprazole intervention was omitted. Other analyses also included interaction terms among the treatment indicators. We performed separate analyses of subjects who were H. pylori–seropositive and –seronegative in 1994 and computed the weighted average of the separately estimated treatment main effects (log odds ratio [OR]) with weights inversely proportional to the stratum-specific main-effect variances. After we tested for null treatment effects and computed 95% confidence intervals on the log odds scale, we exponentiated to produce 95% confidence intervals for the odds ratios. To allow the reader to examine the intervention data in detail, we have presented the prevalences of severity scores of 3 or more and of 6 or more in 1999 and 2003 for each of the eight treatment combinations among subjects who were H. pylori–seropositive in 1994 and for each of the four treatment combinations among subjects who were H. pylori–seronegative in 1994 (Table 1). Categorical distributions were analyzed by omnibus chi-square and stratified Cochran–Armitage–Mantel trend tests. All P values are two-sided, and P<.05 was considered statistically significant. Intention-to-treat analyses were performed on the 3365 eligible subjects. To examine whether the success and/or the duration of H. pylori eradication influenced the histologic severity score in 2003, we used an F test for a nonrandomized comparison among H. pylori–treated subjects in three groups: CUBT-positive in 1996; CUBT-negative in 1996 and CUBT-positive in 2003; and CUBT-negative in 1996 and in 2003. To analyze data on gastric cancer incidence, we used a stratified Cox proportional hazards model to adjust treatment relative hazards for 1994 histopathology strata and for main effects of age, sex, and other treatments. We could not test the proportional hazards assumptions over time because there were too few incident gastric cancers within strata defined by 1994 histopathology.
|
|
RESULTS |
|---|
|
TopNotesAbstractIntroductionPopulation and methodsResultsDiscussionReferences |
|---|
Compliance and H. pylori Eradication
As previously reported (14), pill counts through February 28, 1999, indicated that approximately 95% of study participants took all of their vitamin and garlic supplements. In addition, analyses of blood samples that were collected on a quarterly basis from randomly selected subjects (N = 80) revealed that serum levels of vitamins C and E and of S-allyl cysteine, a marker for garlic supplements, were increased statistically significantly in the active treatment groups compared with the respective placebo groups (14). These compliance patterns continued through March 31, 2003. From November 30, 1995, to May 31, 1996, the vitamin supplement also contained -carotene (7.5 mg twice daily), which was discontinued after May 31, 1996, because of concerns about possible increased lung cancer risk that were raised by results of other trials (15,16). Garlic and vitamin supplements were not given in June and July 1999, and garlic supplements were not given in September 2002 because of interruptions in the availability of the supplements. Histopathology was available from 95.9% of the living eligible subjects in 1999 and from 91.9% of the living eligible subjects in 2003 (Fig. 1).
The initial treatment with amoxicillin and omeprazole successfully eradicated H. pylori infections in 703 (62%) of the 1130 eligible subjects who were H. pylori–seropositive at baseline, as determined by CUBT; the combined eradication rate was 73% (827/1130) after re-treatment with amoxicillin and omeprazole. In 2003, 46% of the amoxicillin/omeprazole group remained free of H. pylori infection, and 10% of the corresponding placebo group were also free from H. pylori infection (as determined by CUBT).
Intervention Effects
From baseline in 1994 to 2003, the percentage of subjects with at least mild dysplasia (i.e., severity score 6) more than doubled, regardless of treatment (Table 2). At baseline, there were no statistically significant differences in the distributions of histopathology categories between the active treatments and their respective placebos. However, in 1999 and 2003 the distribution of histopathology categories in the H. pylori treatment group differed statistically significantly from that in the placebo group (P = .009 and P = .0001, respectively) (Table 2). In particular, compared with the placebo arm, the H. pylori treatment arm had higher proportions of subjects with mild chronic atrophic gastritis in 1999 and 2003 and with superficial gastritis in 2003. There was correspondingly less deep intestinal metaplasia in the H. pylori treatment arm than in the placebo arm in 1999 and 2003. No such differences were seen for the vitamin or garlic treatment arms. Data in Table 2 allow one to gauge the absolute magnitude of intervention effects. For example, in 2003, 69.9% of those treated for H. pylori had severity score 3, compared to 78.4% in the placebo group. For severity score 6, the corresponding results were 35.7% compared to 34.6%.
|
After adjustment for age, baseline histology, sex, and the other interventions, the odds of the combined endpoint of dysplasia or gastric cancer (severity score
6) were not statistically significantly affected by any treatment, either in 1999 or in 2003 (Table 3). H. pylori treatment statistically significantly decreased the odds ratio for the second endpoint, an advanced lesion (i.e., severe chronic atrophic gastritis, intestinal metaplasia, dysplasia, or gastric cancer [severity score 3]) to 0.77 (95% CI = 0.62 to 0.95) in 1999 and to 0.60 (95% CI = 0.47 to 0.75) in 2003. No statistically significant effects on this endpoint were seen for garlic treatment in 1999 or 2003 or for vitamin treatment in 2003. The odds ratio for a severity score of 3 or higher was transiently increased in 1999 in the active vitamin arm to 1.32 (95% CI = 1.12 to 1.57). We examined interactions among interventions for endpoints 1 and 2 in 1999 and 2003, separately among subjects who were seropositive or seronegative for H. pylori in 1994. None of the interactions were consistent or large, and only two of the 16 two-way interactions were nominally statistically significant without adjustment for multiple comparisons (data not shown). Among subjects in the H. pylori intervention group, those who received active treatment had statistically significantly lower average severity scores than those who received placebo in 2003 (4.45 versus 4.69; difference = –0.24 [95% CI = –0.40 to –0.09], P = .002); the average severity score was not reduced by the garlic or vitamin interventions in 1999 or 2003 (Table 3). We analyzed the proportions of the study population for whom the severity score had decreased (indicating disease regression), stayed the same (no change in disease), or increased (indicating disease progression) from 1994 to 1999 and from 1994 to 2003 (Table 4). Compared with placebo, active H. pylori treatment increased the proportion of subjects with disease regression (17% versus 12%) and decreased the proportion of subjects with disease progression (45% versus 49%) in 2003 (Ptrend = .006) but not in 1999 (22% versus 21% and 36% versus 36%, respectively; Ptrend = .63). The garlic intervention had no effect on disease progression or regression from 1994 to either 1999 or 2003. The vitamin intervention had no statistically significant effects on disease progression or regression from 1994 to 2003, but it led to a slight increase in the proportion of subjects with progressing disease in 1999 compared with placebo (38% versus 34%; Ptrend = .043).
|
To investigate whether specific types of precancerous gastric lesions were affected by H. pylori treatment, we examined changes in the distribution of histopathology categories from 1994 to 1999 and to 2003 (Table 5). Among the subjects who had less than severe chronic atrophic gastritis (severity score 0–2) in 1994, 60.0% of those in the active treatment arm and 47.2% of those in the placebo arm remained in this category in 2003; in 1999, 48.3% of subjects in the active arm and 45.7% of subjects in the placebo arm remained in this category (Table 5). Among the subjects who had severe chronic atrophic gastritis in 1994 (severity score 3), 48.0% of those in the active arm and 25.0% of those in the placebo arm had severe chronic atrophic gastritis or a less severe histopathology (severity score 0–3) in 2003; in 1999, 62.5% of subjects in the active arm and 58.0% of subjects in the placebo arm were in this category. Among the subjects who had intestinal metaplasia in 1994, there was little evidence of favorable treatment-induced shifts in histology in 1999 or 2003 (Table 5). Indeed, of the subjects with intestinal metaplasia in 1994, 49.3% of those receiving the active H. pylori treatment and 57.7% of those receiving placebo had intestinal metaplasia or a less severe histopathology (severity score 0–5) in 2003. Among subjects who had dysplasia in 1994, six (4.2%) progressed to gastric cancer with the active H. pylori treatment and 10 (6.6%) progressed to gastric cancer with placebo in 2003; the respective numbers were five (3.3%) and six (3.7%) in 1999. H. pylori treatment also had greater favorable effects on the average severity score (i.e., the difference between the average severity score on the placebo arm minus that on the active arm was greater) in subgroups with less severe baseline histopathology, including nonsmokers, women, and subjects younger than age 45 (data not shown).
|
To examine whether the success and/or the duration of H. pylori eradication would influence treatment effects on average severity score, we conducted a nonrandomized comparison among H. pylori–treated subjects who fell into three categories: CUBT-positive in 1996; CUBT-negative in 1996 and CUBT-positive in 2003; and CUBT-negative in 1996 and in 2003. These three groups had average severity scores in 2003 of 4.71 (95% CI = 3.94 to 4.42), 4.44 (95% CI = 4.24 to 4.64), and 4.18 (95% CI = 4.00 to 4.36), respectively. These average severity scores are ordered as one would expect if successful and durable eradication is more effective than transient eradication in reducing histologic severity. An F test indicated that the differences in severity among these three groups was statistically significant (P = .0016). Thus, more effective H. pylori eradication was associated with a lower average severity score in 2003.
A total of 46 eligible subjects who were H. pylori–seropositive in 1994 were diagnosed with gastric cancer by 2003 (Table 6): 19 (1.68%) of 1130 subjects in the active H. pylori treatment arm and 27 (2.39%) of 1128 subjects in the placebo arm (P = .23). For subjects in each of the 1994 histopathology categories except "missing," those in the active H. pylori treatment group had fewer incident gastric cancers than those in the placebo group (Table 6). The unadjusted relative risk of gastric cancer for these subjects, 0.70 (95% CI = 0.39 to 1.27), was similar to the relative hazards for gastric cancer, 0.64 (95% CI = 0.35 to 1.15; P = .14), which was determined from a Cox proportional hazards model that was stratified on histopathology in 1994 (severity scores of 0–2, 3–4, 5, and 6–8) and adjusted with main effects for age in 1994 (35–39, 40–44, 45–54, and 55 years), sex, and garlic and vitamin treatments. The 58 incident gastric cancers among all eligible subjects were nearly evenly divided between the subjects who were assigned to the vitamin interventions and their placebo control subjects and between those assigned to the garlic interventions and their placebo control subjects (Table 6). The relative hazards for gastric cancer were 1.03 (95% CI = 0.61 to 1.73; P = .91) for the vitamin intervention and 1.06 (95% CI = 0.63 to 1.78; P = .84) for the garlic intervention. These analyses were stratified into 12 categories defined by 1994 histopathology (severity score 0–2, 3, 4–5, and 6–8) and by the combination of 1994 serology for H. pylori with treatment for H. pylori (seronegative for H. pylori, seropositive for H. pylori with amoxicillin/omeprazole treatment, and seropositive for H. pylori with placebo for amoxicillin/omeprazole), and adjusted with main effects for age, sex, and either garlic or vitamin treatment.
|
We observed no statistically significant differences in the numbers of deaths between any of the active treatment groups and their respective placebo groups. In the H. pylori treatment arm, eight subjects died from gastric cancer, 24 from other cancers, and 35 from other causes, compared with 10, 21, and 27 subjects, respectively, in the placebo group. In the vitamin arm, nine subjects died from gastric cancer, 32 from other cancers, and 41 from other causes compared with 12, 30, and 59 subjects, respectively, in the placebo group. In the garlic arm, 12 subjects died from gastric cancer, 31 from other cancers, and 52 from other causes compared with nine, 31, and 48 subjects, respectively, in the placebo group. As previously reported (4,14), apart from rashes seen on the amoxicillin/omeprazole arm, no statistically significant toxicities were noted.
|
DISCUSSION |
|---|
|
TopNotesAbstractIntroductionPopulation and methodsResultsDiscussionReferences |
|---|
The purpose of this trial was to examine whether one-time treatment with amoxicillin/omeprazole or long-term supplementation with vitamin or garlic preparations could impede the development of advanced precancerous gastric lesions in a population with high prevalences of H. pylori infection and precancerous gastric lesions and high gastric cancer mortality rates. Despite excellent treatment compliance, there was no evidence that either vitamin or garlic supplements favorably altered the distribution of precancerous gastric lesions or any of the three main endpoints, namely the proportion of subjects with dysplasia or gastric cancer, the proportion of subjects with severe chronic atrophic gastritis, intestinal metaplasia, dysplasia, or gastric cancer, or the average severity score. In addition, there was no evidence that either vitamin or garlic supplements favorably altered the histopathologic progression or regression rates.
In contrast to the negative findings for vitamin and garlic supplements, H. pylori treatment at the start of the trial reduced the prevalence of severe chronic atrophic gastritis, intestinal metaplasia, dysplasia, and gastric cancer combined and the average severity score for precancerous gastric lesions and induced more regression and less progression of precancerous gastric lesions than placebo. However, H. pylori treatment did not induce statistically significant reductions in the prevalence of dysplasia and gastric cancer combined. The favorable effects on all endpoints were more pronounced at 7.3 years after H. pylori treatment than at 3.3 years after treatment. Favorable changes in the distributions of precancerous gastric lesions were found among subjects who had less than severe chronic atrophic gastritis (severity score 0–2), who had severe chronic atrophic gastritis or less severe histopathology (severity score 0–3), and who had dysplasia (severity score 6–8) in 1994, but not among those who had intestinal metaplasia (severity score 4–5) in 1994. The effects of H. pylori on the average severity score tended to be greater in subgroups with less severe histopathology at baseline, including nonsmokers, women, and subjects younger than 45 years.
Five previous randomized trials of H. pylori treatment have reported mixed results for precancerous gastric lesions. In a 2 x 2 x 2 factorial trial in Narino, Columbia, of H. pylori triple therapy (metronidazole, amoxicillin, and bismuth subsalicylate), vitamin C, and -carotene, Correa et al. (17) found that treatment for 6 years with any combination of the active factors increased the regression rate of precancerous gastric lesions compared with treatment with placebos only. However, our subsequent analysis of data from this trial provided by the authors (18) showed that H. pylori treatment had no statistically significant main effect on the distribution of histopathology categories, on the odds of developing advanced precancerous gastric lesions, on the average severity score, or on the rates of progression or regression of precancerous gastric lesions. All subjects in this trial were offered H. pylori treatment after 6 years on trial (19). Despite the fact that some members of the control group received H. pylori treatment after 6 years, an analysis of main effects at 12 years showed that H. pylori treatment at baseline statistically significantly reduced the histologic severity of precancerous gastric lesions, whereas treatment with -carotene or vitamin C did not (19), in agreement with our findings.
A randomized trial from The Netherlands of 231 H. pylori–positive subjects with gastroesophageal reflux compared H. pylori treatment for 1 week with omeprazole, amoxicillin, and clarithromycin versus treatment with omeprazole alone on changes in gastric histology 2 years later (20). The authors compared subjects in the triple therapy group who were uninfected at the end of the trial with subjects in the omeprazole control group who remained infected at the end of the trial; thus the intention-to-treat principle was not used. These 2-year comparisons indicated statistically significant reductions in inflammation and atrophy but not in intestinal metaplasia in the triple therapy arm compared with omeprazole alone.
Another randomized trial of subjects with CagA-positive strains of H. pylori in Mexico (21) revealed that those treated with omeprazole, amoxicillin, and clarithromycin had a greater decrease in a severity score (defined as a weighted combination of the numbers of sites with various histopathologies) from 6 weeks to 1 year after treatment than those treated with placebo. However, the treatment effect was not statistically significant in a comparison of changes in histology from baseline to 1 year after treatment, and analyses of the proportions with worsening, stable, or improving histology did not differ statistically significantly between the treated and placebo groups, either from 6 weeks to 1 year after treatment or from baseline to 1 year after treatment.
Sung et al. (22) conducted a randomized trial to compare the effects of omeprazole, amoxicillin, and clarithromycin versus placebo on changes in precancerous gastric lesions. Their analyses were restricted to subjects in the active treatment arm who were H. pylori–negative 1 year after treatment and to subjects in the placebo arm who remained H. pylori–positive 1 year after treatment; thus, the intention-to-treat principle was not used. The authors concluded that the anti-Helicobacter therapy did not change either intestinal metaplasia or glandular atrophy. Subsequent 5-year follow-up data from that trial (23) indicated improvement in the degree of gastric atrophy and intestinal metaplasia among the subjects who received H. pylori therapy.
Wong et al. (24) described results of a randomized trial of omeprazole, amoxicillin/clavulanate, and metronidazole versus placebo among subjects in Changle County, Fujian Province, China. In this trial, subjects in the active treatment arm had less progression of gastric atrophy in the antrum and corpus and less progression of intestinal metaplasia in the corpus 5 years after treatment than subjects in the placebo arm.
Thus, the previous randomized controlled trials have yielded mixed results on the effects of H. pylori treatment on precancerous gastric lesions. Two of the three previous reports on effects at 5 or more years after treatment only appeared as abstracts (23,24), and the estimated treatment effects in the other report (19) were potentially attenuated by treatment of the comparison group after 6 years. Nonetheless, these three studies support the strong evidence from the present trial for the long-term effect of one-time H. pylori treatment in reducing the prevalence of advanced precancerous gastric lesions. This result is also supported by several nonrandomized clinical studies that have reported favorable changes in histopathology after successful eradication of H. pylori (25–27). Likewise, in our study, more effective H. pylori eradication was associated with lower average severity scores.
Although this trial was not designed to detect the effects of treatment on gastric cancer incidence, favorable trends were seen. Considering that three previously published randomized trials of treatment for H. pylori reported a total of only 33 gastric cancers, the 46 incident gastric cancers in our study (19 among 1130 subjects in the active H. pylori treatment arm and 27 among 1128 subjects in the placebo arm) add substantially to the evidence from randomized trials that H. pylori eradication may reduce gastric cancer incidence. Correa et al. (17–19) reported three incident gastric cancers among 321 subjects receiving H. pylori antibiotic treatment versus two incident gastric cancers among 309 subjects receiving a placebo during the 6 placebo-controlled years of the study. The largest previous study (28) reported seven gastric cancers among 817 subjects receiving antibiotic treatment versus 11 cancers among 813 subjects receiving a placebo, whereas the third study (29) reported four gastric cancers among 220 subjects receiving antibiotics versus six gastric cancers among 215 subjects receiving a placebo. Pooling the data from all four studies yields ratios of incident gastric cancer cases to populations at risk of 33/2488 or 0.0133 for the active group versus 46/2465 or 0.0187 with a placebo, for a corresponding relative risk of 0.71 (95% CI = 0.45 to 1.11; P = 0.134). This combined relative risk is similar to the relative risks reported in the present trial: 0.70 (unadjusted) and 0.64 (adjusted). The estimate of the protective effect of H. pylori treatment on gastric cancer incidence from the combined data is suggestive but not statistically significant. Despite similar findings from many observational studies (5,30–32), new randomized studies or additional follow-up of ongoing trials will be required to determine whether H. pylori treatment reduces gastric cancer incidence. If such an effect can be shown, additional studies would be needed to evaluate the risks and benefits of programs to screen for and treat H. pylori infection in particular geographic regions. In Linqu County, where there was little toxicity to the omeprazole/amoxicillin regimen (4,14), where the prevalence of H. pylori is 67%, where nearly half of those treated remained free of H. pylori infection for 7.5 years, and where gastric cancer mortality rates are among the highest in the world, it is likely that the benefits would outweigh the risks.
Despite the long duration of and the high compliance with the vitamin intervention, no beneficial effects on the prevalence of precancerous gastric lesions were seen. In fact, the vitamin intervention had transient adverse effects on precancerous gastric lesions at 3.3 years after treatment began, although those effects disappeared by 7.3 years after treatment began. Correa et al. (17) reported favorable effects of vitamin C in combination with other ingredients, but our re-analysis of the 6-year vitamin C data (18) did not reveal statistically significant main effects on the odds of dysplasia or gastric cancer combined or on the odds of intestinal metaplasia, dysplasia, or gastric cancer combined; moreover, a subsequent analysis after 12 years of follow-up (19) revealed no long-term main effect on histologic severity. We are unaware of any other randomized trial that has evaluated the effect of vitamin C on the prevalence of precancerous gastric lesions. However, data from observational studies have suggested that vitamin C is associated with reduced prevalence of intestinal metaplasia (9) and reduced progression to dysplasia or gastric cancer (10). A meta-analysis of randomized trials that examined the effect of vitamin supplements on gastric cancer incidence or mortality found no evidence that either vitamin C in combination with -carotene and/or vitamin E nor vitamin E alone can prevent gastric cancer (33). In a trial in Henan Province, China (11), a combination of vitamin E, selenium, and -carotene was associated with reduced gastric cancer mortality, whereas supplementation with a combination of vitamin C and molybdenum was not.
To our knowledge, no other randomized trials of the effect of long-term garlic supplementation on the prevalence of precancerous gastric lesions has been reported. We found no effect for a combination of aged garlic extract with steam-distilled garlic oil. A randomized trial (34) in Qixia County, Shandong Province, China, compared the effect of a combination of the garlic-derived organosulfide diallyl trisulfide ("allitridum") and sodium selenite against placebo, each given for 1 month each year for 3 consecutive years. Incident gastric cancers were found in 23 of 2526 subjects given the active intervention, compared with 30 of 2507 subjects receiving a placebo (P = .32).
This study has several potential limitations. First, one can question some of the choices made in designing and analyzing this trial, such as the agents and doses studied. For example, other garlic preparations, such as garlic powder or raw or cooked garlic, or other vitamin supplements or doses might have been effective. Second, variability in biopsy sampling and reading of the histopathology may have reduced the power to detect intervention effects of garlic and vitamin supplementation. However, the decisive results for amoxicillin/omeprazole treatment, which were based on smaller samples than the garlic and vitamin comparisons, demonstrate that our methods for histopathologic classification are informative. We might have chosen other ways to combine and analyze the histopathologic data from the various biopsy sites. The present analyses were based on prespecified endpoints set out in the protocol, but further analyses based on other features of the histopathology may be useful.
In summary, we conducted a large, long-term randomized trial that included most of the adults aged 35–64 years in 13 randomly selected villages in Linqu County. One-time H. pylori treatment with amoxicillin and omeprazole induced statistically significant reductions in the combined prevalence of severe chronic atrophic gastritis, intestinal metaplasia, dysplasia, or gastric cancer, in histologic severity, and in the progression of precancerous gastric lesions, but not in the combined prevalence of dysplasia or gastric cancer. There were also favorable but non–statistically significant trends in gastric cancer incidence with H. pylori treatment. Supplementation for 7.3 years with vitamin or garlic preparations had no such effects. Thus, a safe and simple H. pylori treatment retarded the progression of precancerous gastric lesions. We are continuing to follow the study population to obtain information on gastric cancer incidence and cause-specific mortality, and we are offering annual endoscopic examinations to those with moderate or severe dysplasia or any dysplasia in two or more biopsy sites.
|
NOTES |
|---|
|
TopNotesAbstractIntroductionPopulation and methodsResultsDiscussionReferences |
|---|
This research was supported by the Intramural Research Program of the National Institutes of Health, National Cancer Institute, and in part by National Cancer Institute Contracts NO2-CP-71103 and NO2-CP-21169. National Cancer Institute contracts also supported work at Westat and at Information Management Services. Drs. Mitchell H. Gail and Wei-cheng You had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. The study sponsors did not directly participate in developing the protocol or analyzing or interpreting the data or writing the manuscript or the decision to submit the manuscript for publication.
We are grateful to the residents, field staff, and government of Linqu County for supporting this long-term trial. We thank Dr. Haru Amagase and others at Wakunaga of America, Ltd. for providing technical advice and assays of S-allyl cysteine and garlic supplements; Shanghai Squibb for providing vitamin supplements; and Astra Corporation for providing amoxicillin and omeprazole. We thank Drs. Z. X. Han, B. Q. Yang, Y. Li, and L. Shen for performing endoscopic examinations, Drs. S. Hu and Y. Q. Xie for histopathologic analyses, Drs. P. Correa and P. Sipponen for pathology quality-control studies, Dr. C. S. Yang for vitamin assays, Dr. J. Jiang for urea breath testing, Dr. Steven D. Mark for helpful discussions on design, and Linda Lannom, John P. Heinrich, Duke Owen, Yizhu Chen, and data managers and other programmers at Westat for administrative, logistical, and data management support. We also thank members of the Data Safety and Monitoring Committee for guidance and oversight throughout the study. The Committee included Mimi Yu (Chair 2000–2004), Junshi Chen, David Fleischer, David Graham, Stephen Piantadosi, and Alison Wichman.
|
REFERENCES |
|---|
|
TopNotesAbstractIntroductionPopulation and methodsResultsDiscussionReferences |
|---|
(1) Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005;55:74–108.
(2) Wang TG, You WC, Henderson BE, Blot WJ. A case–control study of stomach cancer in Shandong Province. J Natl Cancer Inst Monogr 1985;69:9–10.[Medline]
(3) You WC, Blot WJ, Li JY, Chang YS, Jin ML, Kneller R, et al. Precancerous gastric lesions in a population at high risk of stomach cancer. Cancer Res 1993;53:1317–21.
(4) Gail MH, You WC, Chang YS, Zhang L, Blot WJ, Brown LM, et al. Factorial trial of three interventions to reduce the progression of precancerous gastric lesions in Shandong, China: design issues and initial data. Control Clin Trials 1998;19:352–69.[CrossRef][ISI][Medline]
(5) International Agency for Research on Cancer. Monographs on the evaluation of carcinogenic risks to humans. Vol 61: Schistosomes, liver flukes and Helicobacter pylori. Lyon (France): International Agency for Research on Cancer; 1994. p. 177–240.
(6) You WC, Li JY, Blot WJ, Chang YS, Jin ML, Gail MH, et al. Evolution of precancerous lesions in a rural Chinese population at high risk of gastric cancer. Int J Cancer 1999;83:615–9.[CrossRef][ISI][Medline]
(7) Correa P. The biological model of gastric carcinogenesis. IARC Sci Publ 2004;157:301–10.
(8) You WC, Blot WJ, Chang YS, Ershow AG, Yang ZT, An Q, et al. Diet and high risk of stomach cancer in Shandong, China. Cancer Res 1988;48:3518–23.
(9) Zhang L, Blot WJ, You WC, Chang YS, Liu XQ, Kneller RW, et al. Serum micronutrients in relation to pre-cancerous gastric lesions. Int J Cancer 1994;56:650–4.[ISI][Medline]
(10) You WC, Zhang L, Gail MH, Chang YS, Liu WD, Ma JL, et al. Gastric dysplasia and gastric cancer: Helicobacter pylori, serum vitamin C, and other risk factors. J Natl Cancer Inst 2000;92:1607–12.
(11) Blot WJ, Li JY, Taylor PR, Guo W, Dawsey S, Wang GQ, et al. Nutrition intervention trials in Linxian, China: supplementation with specific vitamin/mineral combinations, cancer incidence, and disease-specific mortality in the general population. J Natl Cancer Inst 1993;85:1483–92.
(12) You WC, Blot WJ, Chang YS, Ershow AG, Yang ZT, An Q, et al. Allium vegetables and reduced risk of stomach cancer. J Natl Cancer Inst 1989;81:162–4.
(13) Dixon MF, Genta RM, Yardley JH, Correa P. Classification and grading of gastritis: the updated Sydney System. International Workshop on the Histopathology of Gastritis, Houston 1994. Am J Surg Pathol 1996;20:1161–81.[CrossRef][ISI][Medline]
(14) You WC, Chang YS, Heinrich J, Ma JL, Liu WD, Zhang L, et al. An intervention trial to inhibit the progression of precancerous gastric lesions: compliance, serum micronutrients and S-allyl cysteine levels, and toxicity. Eur J Cancer Prev 2001;10:257–63.[CrossRef][ISI][Medline]
(15) The effect of vitamin E and beta carotene on the incidence of lung cancer and other cancers in male smokers: the Alpha-Tocopherol, Beta Carotene Cancer Prevention Study Group. N Engl J Med 1994;330:1029–35.
(16) Omenn GS, Goodman GE, Thornquist MD, Balmes J, Cullen MR, Glass A, et al. Effects of a combination of -carotene and vitamin A on lung cancer and cardiovascular disease. N Engl J Med 1996;334:1150–5.
(17) Correa P, Fontham ET, Bravo JC, Bravo LE, Ruiz B, Zarama G, et al. Chemoprevention of gastric dysplasia: randomized trial of antioxidant supplements and anti-Helicobacter pylori therapy. J Natl Cancer Inst 2000;92:1881–8.
(18) Correa P, Fontham ET, Bravo JC, Bravo LE, Ruiz B, Zarama G, et al. Re: Chemoprevention of gastric dysplasia: randomized trial of antioxidant supplements and anti-Helicobacter pylori therapy (letter). J Natl Cancer Inst 2001;93:559.
(19) Mera R, Fontham ET, Bravo LE, Bravo JC, Piazuelo MB, Camargo MC, et al. Long term follow up of patients treated for Helicobacter pylori infection. Gut 2005;54:1536–40.
(20) Kuipers EJ, Nelis GF, Klinkenberg-Knol EC, Snel P, Goldfain D, Kolkman JJ, et al. Cure of Helicobacter pylori infection in patients with reflux oesophagitis treated with long term omeprazole reverses gastritis without exacerbation of reflux disease: results of a randomised controlled trial. Gut 2004;53:12–20.
(21) Ley C, Mohar A, Guarner J, Herrera-Goepfert R, Figueroa LS, Halperin D, et al. Helicobacter pylori eradication and gastric preneoplastic conditions: a randomized, double-blind, placebo-controlled trial. Cancer Epidemiol Biomarkers Prev 2004;13:4–10.
(22) Sung JJ, Lin SR, Ching JY, Zhou LY, To KF, Wang RT, et al. Atrophy and intestinal metaplasia one year after cure of H. pylori infection: a prospective, randomized study. Gastroenterology 2000;119:7–14.[CrossRef][ISI][Medline]
(23) Sung JJ, Lin SR, Leung W, Ng EK, Ching JY, To KF, et al. Does eradication of H. pylori prevent deterioration of gastric atrophy and intestinal metaplasia? A 5-year follow-up. Gastroenterology 2002;122:A170.
(24) Wong BCY, Lam SK, Wong WM, Feng R, Zheng TT, Yuen ST, et al. Eradication of Helicobacter pylori infection significantly slows down the progression of precancerous lesions in high risk population: a 5-year prospective randomized study. Gastroenterology 2002;122:A588.
(25) Ohkusa T, Fujiki K, Takashimizu I, Kumagai J, Tanizawa T, Eishi Y, et al. Improvement in atrophic gastritis and intestinal metaplasia in patients in whom Helicobacter pylori was eradicated. Ann Intern Med 2001;134:380–6.
(26) Kokkola A, Sipponen P, Rautelin H, Harkonen M, Kosunen TU, Haapiainen R, et al. The effect of Helicobacter pylori eradication on the natural course of atrophic gastritis with dysplasia. Aliment Pharmacol Ther 2002;16:515–20.[CrossRef][ISI][Medline]
(27) Ito M, Haruma K, Kamada T, Mihara M, Kim S, Kitadai Y, et al. Helicobacter pylori eradication therapy improves atrophic gastritis and intestinal metaplasia: a 5-year prospective study of patients with atrophic gastritis. Aliment Pharmacol Ther 2002;16:1449–56.[CrossRef][ISI][Medline]
(28) Wong BC, Lam SK, Wong WM, Chen JS, Zheng TT, Feng RE, et al. Helicobacter pylori eradication to prevent gastric cancer in a high-risk region of China: a randomized controlled trial. JAMA 2004;291:187–94.
(29) Leung WK, Lin SR, Ching JY, To KF, Ng EK, Chan FK, et al. Factors predicting progression of gastric intestinal metaplasia: results of a randomised trial on Helicobacter pylori eradication. Gut 2004;53:1244–9.
(30) Helicobacter and Cancer Collaborative Group. Gastric cancer and Helicobacter pylori: a combined analysis of 12 case control studies nested within prospective cohorts. Gut 2001;49:347–53.
(31) Uemura N, Okamoto S, Yamamoto S, Matsumura N, Yamaguchi S, Yamakido M, et al. Helicobacter pylori infection and the development of gastric cancer. N Engl J Med 2001;345:784–9.
(32) Ohata H, Kitauchi S, Yoshimura N, Mugitani K, Iwane M, Nakamura H, et al. Progression of chronic atrophic gastritis associated with Helicobacter pylori infection increases risk of gastric cancer. Int J Cancer 2004;109:138–43.[CrossRef][ISI][Medline]
(33) Bjelakovic G, Nikolova D, Simonetti RG, Gluud C. Antioxidant supplements for prevention of gastrointestinal cancers: a systematic review and meta-analysis. Lancet 2004;364:1219–28.[CrossRef][ISI][Medline]
(34) Li H, Li HQ, Wang Y, Xu HX, Fan WT, Wang ML, et al. An intervention study to prevent gastric cancer by micro-selenium and large dose of allitridum. Chin Med J 2004;117:1155–60.[Medline]
Manuscript received December 3, 2005; revised May 9, 2006; accepted May 23, 2006.
Correspondence about this Article
Editorial about this Article
CiteULike 
Connotea 
Del.icio.us What's this?
J Natl Cancer Inst 2006 98: 1426.
J Natl Cancer Inst 2006 98: 945-946.
This article has been cited by other articles:
|
|
 |
|
  M. J. Blaser Understanding Microbe-Induced Cancers Cancer Prevention Research, June 1, 2008; 1(1): 15 - 20. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Yeh, K. M. Kuntz, M. Ezzati, C. Hur, C. Y. Kong, and S. J. Goldie Development of an Empirically Calibrated Model of Gastric Cancer in Two High-Risk Countries Cancer Epidemiol. Biomarkers Prev., May 1, 2008; 17(5): 1179 - 1187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.-W. Lee, B. Rickman, A. B. Rogers, Z. Ge, T. C. Wang, and J. G. Fox Helicobacter pylori Eradication Prevents Progression of Gastric Cancer in Hypergastrinemic INS-GAS Mice Cancer Res., May 1, 2008; 68(9): 3540 - 3548. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Hamashima, D. Shibuya, H. Yamazaki, K. Inoue, A. Fukao, H. Saito, and T. Sobue The Japanese Guidelines for Gastric Cancer Screening Jpn. J. Clin. Oncol., April 1, 2008; 38(4): 259 - 267. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. R. Taylor Prevention of Gastric Cancer: A Miss J Natl Cancer Inst, January 17, 2007; 99(2): 101 - 103. [Full Text] [PDF]
|
|
|
|
|
|
M. Plummer, J. Vivas, G. Lopez, J. C. Bravo, S. Peraza, E. Carillo, E. Cano, D. Castro, O. Andrade, V. Sanchez, et al. Chemoprevention of Precancerous Gastric Lesions With Antioxidant Vitamin Supplementation: A Randomized Trial in a High-Risk Population J Natl Cancer Inst, January 17, 2007; 99(2): 137 - 146. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 JournalScan Gut, November 1, 2006; 55(11): 1639 - 1639. |