• gastric cancer
• Helicobacter pylori

We would like to comment on the relationship of chronic Helicobacter pylori infection and gastric cardia cancer risk discussed in the recent combined analysis of prospective serological studies presented by the Helicobacter and Cancer Collaborative Group (Gut 2001;49:347–53). It is our opinion that the null association between H pylori seropositivity and gastric cardia cancer risk reported in this analysis may have resulted in a large part from the combination of observational studies with (a) different definitions of gastric cardia cancer and (b) diverse subject populations, particularly with respect to the prevalence of gastro-oesophageal reflux disease (GORD), which we believe may affect the perceived H pylori-cardia cancer relationship.

By definition, the gastric cardia represents only the proximal 2–3 cm of the stomach. This small anatomical region can easily be overgrown by tumours that originate from adjacent mucosal sites. Thus in populations where both oesophageal and gastric adenocarcinomas are common, tumours described as “cardia cancers” undoubtedly include a mixture of neoplasms arising from the lower oesophagus as well as the gastric cardia (and perhaps even the more distal stomach). Classification of gastric cardia cancer has been further confused by the inclusion of malignant lesions situated 2–3 cm above the gastro-oesophageal junction in some previous reports.¹

With respect to disease aetiology, adenocarcinoma of the oesophagus appears to be strongly associated with GORD,¹ while adenocarcinoma of the distal stomach has been positively linked with chronic H pylori carriage (Gut 2001;49:347–53). It is therefore not surprising that both GORD and H pylori have been suggested as potential risk factors for “gastric cardia cancer”. In our opinion, GORD is probably associated with most oesophageal adenocarcinomas and some true gastric cardia cancers whereas chronic H pylori exposure appears to be a predisposing factor for most gastric non-cardia adenocarcinomas and some gastric cardia cancers. Additionally, chronic H pylori carriage is thought to reduce the risk of GORD associated tumours by lowering the acid content of refluxed gastric contents² while GORD is not thought to affect the risk of H pylori associated tumours.

Since 1985, our group has collaborated on multiple studies of oesophageal and gastric cancer in Linxian, China. Residents of this region are known to be at high risk for oesophageal squamous cell carcinoma and gastric cardia cancer, moderate risk for gastric non-cardia adenocarcinomas, and very low risk for oesophageal adenocarcinoma. The H pylori seropositivity rate among cancer free adults has been reported as 56%.³ Based on the endoscopic evaluation of nearly 7000 asymptomatic adults in our studies in Linxian, reflux oesophagitis is uncommon in this population. Moreover, no cases of Barrett's oesophagus or oesophageal adenocarcinoma have been histologically confirmed. In contrast, we have detected more than 175 cases of early gastric cardia adenocarcinoma, almost all of which were confined to the 2–3 cm region below the squamocolumnar junction at the time of diagnosis. Other groups have also reported that reflux oesophagitis and its complications are uncommon and that the H pylori carriage rates are high among Asian adults.^4–6 Therefore, such populations seem well suited to the investigation of H pylori and gastric cardia cancer risk, as the potential disease modifying effects of GORD are virtually non-existent. In the combined analysis from the Helicobacter and Cancer Collaborative Group (Gut 2001;49:347–53), our analysis of the pooled data from the three Asian studies revealed a summary odds ratio estimate of 1.67 (95% confidence interval (CI) 1.06–2.64) for H pylori seropositivity and gastric cardia cancer risk. Interestingly, each of the Asian studies further reported similar risk estimates for gastric cardia and gastric non-cardia cancers among H pylori positive subjects, which suggests that the bacterium's putatively procarcinogenic effects may be uniform throughout the stomach.

In most Western populations, GORD is a relatively common disorder,^1,7 while H pylori infection is on the decline.^2,8 In this setting, estimating the risk of H pylori associated gastric cardia cancer becomes substantially more challenging. Our further analysis of the combined data from the seven western studies alone yielded a summary odds ratio of 0.60 (95% CI 0.38–0.93) for H pylori seropositivity and gastric cardia cancer risk. Because the proportion of oesophageal, gastric cardia, and gastric non-cardia cancers included in these studies cannot be readily determined, we believe that interpreting this risk estimate as showing a protective (or null) association is problematic. Consistent with the discussion offered above, the apparent lack of association between H pylori exposure and gastric cardia cancer in Western populations may be due to an over representation of misclassified GORD associated lower oesophageal malignancies in these studies. The appropriateness of combining the existing Western and Asian studies of H pylori and gastric cardia cancer risk is further called into question by formal statistical testing. When we do a heterogeneity test that allows for potential differences between Western and Asian studies, we find that the odds ratios are significantly different (p=0.002).

In summary, we believe that H pylori carriage is a risk factor for adenocarcinoma throughout the stomach, including the gastric cardia. The different conclusion reached by the recently published combined analysis seems likely to have been influenced by pooling data from subject populations with demonstrated or plausible differences in disease classification and disease aetiology, respectively. More specifically, we believe it is difficult to accurately judge the relationship between H pylori seropositivity and gastric cardia cancer risk among populations wherein tumour location has not been rigorously defined and GORD is highly prevalent (due to the potentially misleading biological effects associated with this condition). Thus we respectfully disagree with the conclusion proposed by the Helicobacter and Cancer Collaborative Group that “H pylori does not appear to increase the risk of cardia cancer”. However, we do agree that additional prospective studies, with larger case numbers and longer follow up intervals, would be useful in clarifying this issue.