Justification for Use of HP Eradication in the Prevention of GC

The most accepted model of gastric carcinogenesis is that GC, like colonic cancers, progresses through a cancer cascade. This is particularly true for intestinal-type GC, where evidence shows there is a multistep progression from gastritis to glandular atrophy, intestinal metaplasia (IM), dysplasia, and ultimately cancer. In early observational studies of the natural history of GC, it seemed that an unknown inciting agent triggered the start of this carcinogenesis cascade, which ultimately resulted in cancer. Since 1982, when Marshall and Warren identified HP and causally linked the bacterium with chronic gastritis and peptic ulcer disease, it has become evident that the putative pathogenic agent causing chronic gastritis in the majority of cases is HP. The hypothesis that prevention of GC may be achieved through HP eradication is based on the results of large-scale epidemiologic studies, experimental models, and meta-analysis of case-control studies.

In 1991, 3 seminal prospective case-control studies involving 324 GC subjects with matched controls were published, which demonstrated that the odds ratio (OR) for HP in subjects with GC was between 2.77 and 6.0, with the pooled analysis of the 3 prospective epidemiologic studies demonstrating a relative risk (RR) of 3.8, which was significant. This led to the 1994 classification by the International Agency for Research on Cancer/World Health Organization of HP as a definite carcinogen based on the inference that HP causes histologic gastritis, which, when prolonged, may induce atrophic gastritis, considered the first step in the gastritis-metaplasia-carcinoma sequence of the stomach. Unfortunately, subsequent studies found conflicting results, which prompted several meta-analyses to try to resolve the issue. Of these, 2 meta-analyses used more rigorous entry criteria, including only prospective case-control studies, and both demonstrated pooled ORs for gastric cancer in HP-positive individuals of between 2 and 3. So, in summary, several meta-analyses have found that HP infection is moderately to strongly associated with the risk of GC development. Coupled with the long lead times presumed involved in the transition between stages in the GC cascade, observational studies showing that nonatrophic gastritis progresses to glandular atrophy and IM over a period of 12 years, make HP eradication a natural choice for chemopreventive strategies, with a large window of opportunity during which to institute therapy. However, not all individuals infected with HP develop GC and the reason why certain individuals have a greater propensity to develop GC is most certainly a multi factorial process. So, although it may seem biologically plausible that HP eradication would result in a reduction in GC incidence, given these complex interactions between host factors, HP infection, and environmental factors, such as diet, it is critical that the evidence for the efficacy of HP eradication in GC prevention is examined.

Evidence for Effectiveness of HP Eradication on Gastric Cancer Incidence

Interventional studies of HP eradication would confirm whether eradication of this infection could prevent GC. Large population-based studies have not been performed, however, and are difficult to undertake for the following reasons. First, the incidence of GC is low, and second, the natural history of GC development is long, so that some investigators have projected that more than 35,000 subjects need to be enrolled and followed for more than 10 years to demonstrate a 50% reduction in the GC incidence after HP eradication. This represents a possibly insurmountable issue in the design, execution, and financing of any proposed study, and these problems are demonstrated in the RCTs performed to examine this issue (discussed later). Finally, there are ethical issues involved in randomizing subjects from a high-risk GC region to placebo when there is strong evidence of HP’s putative pathologic role in the development of GC. For these reasons, the authors believe both major cohort studies as well as all RCTs examining the efficacy of HP eradication on GC incidence warrant discussion. Table 1 details the 6 major cohort studies examining the efficacy of HP eradication on GC incidence. Three of the studies are retrospective studies, which included more than 80,000 subjects and found that HP eradication was consistently and significantly associated with reduced GC incidence. The largest of these studies was conducted in Taiwan and demonstrated that in patients with peptic ulcer disease, HP eradication within an arbitrarily defined time frame (defined by the investigators as within a year of hospitalization for peptic ulcer disease) was associated with GC incidence rates similar to those of the general population. Moreover, this effect seemed time dependent, so that as the years posteradication of HP increased, this had the effect of decreasing the GC standardized incidence ratios (SIRs). The remaining 3 cohort studies were prospective interventional studies, which again found benefit for HP eradication in reducing GC incidence. A study by Uemura and colleagues is important because it was one of the first prospective studies to demonstrate that HP eradication caused a reduction in GC incidence and the subjects with duodenal ulcers seemed protected from GC development. The study’s major flaws, however, are the significantly reduced follow-up time in the HP-eradicated cohort (mean 4.8 vs 8.5 years, respectively; P <.001) compared with the untreated cohort, meaning that GC developing later may have been missed in the HP-eradicated/negative group. Moreover, this study was never intended as an interventional study. The study by Yanaoka and colleagues used pepsinogen levels as a surrogate marker of gastric atrophy, which probably would be more cost effective and less invasive if used in a GC chemoprevention program. The investigators suggest that there seems to be a point of no return in the stomach carcinogenesis cascade when HP eradication is of limited benefit. Similarly, the study by Take and colleagues suggests HP eradication was most efficacious before the onset of significant atrophy. They noted, however, that HP eradication even in patients with mild gastric atrophy did not completely prevent GC.

The most robust evidence for the efficacy of HP eradication on GC incidence is derived from RCTs. To date, 5 population-based RCTs have been conducted investigating the effect ( Table 2 ) after excluding earlier publications reporting outcomes from the same populations. A further study from Japan of a similar design examined the risk of GC post–HP eradication in subjects with treated early GC is also included. A recent metanalysis was performed to answer the question of whether HP eradication reduced the GC incidence and included the 6 RCTs listed in Table 2 , although the investigators used some of the data from earlier publications of the same trials due to some changes in methodology in the latest iterations. The meta-analysis found that there was no reduction in the incidence of GC in subjects allocated to HP eradication (RR 0.65; 95% CI, 0.42–1.01) versus placebo (RR 0.70; 95% CI, 0.46–1.08). Unfortunately, there were some serious questions raised about 1 of the data sets included in the meta-analyses due to a reduction in the numbers of GC reported at the 10-year follow-up compared with the 5-year follow-up. Furthermore, 2 of the studies were published in abstract form only, limiting the rigorous assessment of the study methods, analyses, and results, although for the RCT by Zhou and colleagues, information could be obtained from earlier reports of the same data set. In light of the questions raised by the inclusion of these studies in this meta-analysis, in depth examination of the individual studies is warranted.

Only 2 studies have followed patients up for longer than 10 years, although both did not design their studies to detect a difference in GC incidence as a primary outcome and both studies demonstrated a protocol change during the follow-up period. The study by Mera and colleagues followed HP-positive patients for 12 years and did not find a difference in the GC incidence between HP-eradicated and HP-uneradicated cohorts but did find that successful and long-term HP eradication induced more regression and less progression on precancerous lesions. Moreover, this study found that HP eradication was more efficacious on less-advanced lesions and correlated exponentially with the number of years free of HP infection. The investigators postulated that the effects of HP eradication on GC incidence may not be evident in the first 3 to 6 years of follow-up due to the prolonged time required for healing of precancerous lesions. The study by Ma and colleagues was a factorial design study that was originally designed to detect the effect of HP eradication and/or supplementation with garlic extract/vitamins. This study did find a reduction in the incidence of GC at 14.7 years’ follow-up (OR 0.61; 95% CI, 0.38–0.96), which was not seen at the 7.3 years’ follow-up. The findings of this study seem to support the observation that HP eradication’s effects may not be seen in the first 3 to 6 years of follow-up and may explain the results of the only meta-analyses performed to address this issue because half of the included RCTs had follow-up shorter than 10 years. These studies also demonstrate the difficulty with the logistics of executing a RCT to examine this issue. In the two studies with more than 10 year’s follow up, there was a protocol change, including un-blinding and relaxation of endoscopic follow up, which may impact the findings of the study.

The publication by Wong and colleagues remains to date the only population-based RCT where the presented primary outcome is GC incidence post–HP eradication. This study found no difference in GC incidence at 7.5 years. The post hoc analyses found that HP eradication prevented GC development in the cohort without precancerous lesions at time of treatment. This finding again suggests that there is a point of no return, whereby HP eradication fails to reduce GC development, which carries significant biologic plausibility. Finally, the study by Fukase and colleagues, although included in the meta-analysis by Fuccio and colleagues, studied a completely different population from the other RCTs. This study examined patients with early GC who were HP positive and were randomized to HP eradication or to a control group and found that the OR for developing a metachronous GC was 3.53 (95% CI, 0.161–0.775; P = .009). This study raises several issues. First, the population studied already had GC, albeit cured endoscopically, and the findings should not be extrapolated to subjects with high GC risk and HP positivity. Second, from what is known about HP’s role in the development of GC, randomization of patients with GC to a control arm raises ethical issues. Finally, the best available evidence from large cohort studies and RCTs suggests there is a point of no return in terms of GC prevention and HP eradication, when HP eradication seems less efficacious. This study seems to suggest that even in patients who have developed early GC, HP eradication is still of significant benefit despite the gastric histology being past the point of no return. In summary, the Asia-Pacific region and South America have some of the highest rates of HP infection and GC prevalence rates, which explains why all the major cohort studies and RCTs on the effect of HP eradication on GC development have emerged from these regions. A meta-analysis has found that HP eradication does not prevent the development of GC, however questions about some of the studies included in the data set may impact on this conclusion. Examination of individual RCTs examining this issue suggest that HP eradication seems to prevent the development of GC in high-risk populations if instituted prior to the development of significant precancerous lesions. RCTs with GC incidence outcomes as the primary outcome and with longer-term follow-up are needed, however, before chemoprevention with HP eradication can be recommended in areas of high GC prevalence. In subjects with early gastric cancers, the evidence seems to suggest that HP eradication is still of significant benefit.

Justification for Use of HP Eradication in the Prevention of GC

The most accepted model of gastric carcinogenesis is that GC, like colonic cancers, progresses through a cancer cascade. This is particularly true for intestinal-type GC, where evidence shows there is a multistep progression from gastritis to glandular atrophy, intestinal metaplasia (IM), dysplasia, and ultimately cancer. In early observational studies of the natural history of GC, it seemed that an unknown inciting agent triggered the start of this carcinogenesis cascade, which ultimately resulted in cancer. Since 1982, when Marshall and Warren identified HP and causally linked the bacterium with chronic gastritis and peptic ulcer disease, it has become evident that the putative pathogenic agent causing chronic gastritis in the majority of cases is HP. The hypothesis that prevention of GC may be achieved through HP eradication is based on the results of large-scale epidemiologic studies, experimental models, and meta-analysis of case-control studies.

In 1991, 3 seminal prospective case-control studies involving 324 GC subjects with matched controls were published, which demonstrated that the odds ratio (OR) for HP in subjects with GC was between 2.77 and 6.0, with the pooled analysis of the 3 prospective epidemiologic studies demonstrating a relative risk (RR) of 3.8, which was significant. This led to the 1994 classification by the International Agency for Research on Cancer/World Health Organization of HP as a definite carcinogen based on the inference that HP causes histologic gastritis, which, when prolonged, may induce atrophic gastritis, considered the first step in the gastritis-metaplasia-carcinoma sequence of the stomach. Unfortunately, subsequent studies found conflicting results, which prompted several meta-analyses to try to resolve the issue. Of these, 2 meta-analyses used more rigorous entry criteria, including only prospective case-control studies, and both demonstrated pooled ORs for gastric cancer in HP-positive individuals of between 2 and 3. So, in summary, several meta-analyses have found that HP infection is moderately to strongly associated with the risk of GC development. Coupled with the long lead times presumed involved in the transition between stages in the GC cascade, observational studies showing that nonatrophic gastritis progresses to glandular atrophy and IM over a period of 12 years, make HP eradication a natural choice for chemopreventive strategies, with a large window of opportunity during which to institute therapy. However, not all individuals infected with HP develop GC and the reason why certain individuals have a greater propensity to develop GC is most certainly a multi factorial process. So, although it may seem biologically plausible that HP eradication would result in a reduction in GC incidence, given these complex interactions between host factors, HP infection, and environmental factors, such as diet, it is critical that the evidence for the efficacy of HP eradication in GC prevention is examined.

Evidence for Effectiveness of HP Eradication on Gastric Cancer Incidence

Interventional studies of HP eradication would confirm whether eradication of this infection could prevent GC. Large population-based studies have not been performed, however, and are difficult to undertake for the following reasons. First, the incidence of GC is low, and second, the natural history of GC development is long, so that some investigators have projected that more than 35,000 subjects need to be enrolled and followed for more than 10 years to demonstrate a 50% reduction in the GC incidence after HP eradication. This represents a possibly insurmountable issue in the design, execution, and financing of any proposed study, and these problems are demonstrated in the RCTs performed to examine this issue (discussed later). Finally, there are ethical issues involved in randomizing subjects from a high-risk GC region to placebo when there is strong evidence of HP’s putative pathologic role in the development of GC. For these reasons, the authors believe both major cohort studies as well as all RCTs examining the efficacy of HP eradication on GC incidence warrant discussion. Table 1 details the 6 major cohort studies examining the efficacy of HP eradication on GC incidence. Three of the studies are retrospective studies, which included more than 80,000 subjects and found that HP eradication was consistently and significantly associated with reduced GC incidence. The largest of these studies was conducted in Taiwan and demonstrated that in patients with peptic ulcer disease, HP eradication within an arbitrarily defined time frame (defined by the investigators as within a year of hospitalization for peptic ulcer disease) was associated with GC incidence rates similar to those of the general population. Moreover, this effect seemed time dependent, so that as the years posteradication of HP increased, this had the effect of decreasing the GC standardized incidence ratios (SIRs). The remaining 3 cohort studies were prospective interventional studies, which again found benefit for HP eradication in reducing GC incidence. A study by Uemura and colleagues is important because it was one of the first prospective studies to demonstrate that HP eradication caused a reduction in GC incidence and the subjects with duodenal ulcers seemed protected from GC development. The study’s major flaws, however, are the significantly reduced follow-up time in the HP-eradicated cohort (mean 4.8 vs 8.5 years, respectively; P <.001) compared with the untreated cohort, meaning that GC developing later may have been missed in the HP-eradicated/negative group. Moreover, this study was never intended as an interventional study. The study by Yanaoka and colleagues used pepsinogen levels as a surrogate marker of gastric atrophy, which probably would be more cost effective and less invasive if used in a GC chemoprevention program. The investigators suggest that there seems to be a point of no return in the stomach carcinogenesis cascade when HP eradication is of limited benefit. Similarly, the study by Take and colleagues suggests HP eradication was most efficacious before the onset of significant atrophy. They noted, however, that HP eradication even in patients with mild gastric atrophy did not completely prevent GC.

Table 1

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related posts:

Gastric Cancer Screening and Treating Intermediate Lesions to Prevent Gastric Cancer Gastric Cancer Gastric Cancer Surgical Considerations in the Treatment of Gastric Cancer Gastric Mucosal-Associated Lymphoid Tissue Lymphoma

Stay updated, free articles. Join our Telegram channel

Join