Statistics from Altmetric.com
We read with interest the report by Doorakkers et al describing the effect of Helicobacter pylori eradication in preventing the development of non-cardia gastric adenocarcinoma in a population-based Swedish cohort.1 The authors used nationwide Swedish registries to investigate the incidence of gastric adenocarcinoma after H. pylori eradication treatment. Specifically, standardised incidence ratios (SIRs) were used to examine the incidence of gastric adenocarcinoma in individuals treated for H. pylori compared with a background population of corresponding age, sex and calendar year distribution. The time trend of the SIRs after H. pylori eradication was assessed for up to 7.5 years.
The results showed that the risk of non-cardia gastric adenocarcinoma sharply decreased over time after eradication treatment, reaching levels below those of the corresponding background population. The SIRs were 10.74 (95% CI 7.77 to 14.46) at 1–3 years, 2.67 (95% CI 1.63 to 4.13) at 3–5 years and 0.43 (95% CI 0.16 to 0.93) at 5–7.5 years after treatment. The authors concluded that eradication treatment for H. pylori may prevent the development of non-cardia gastric adenocarcinoma.1
However, we suspect the observed trend in the SIRs can be attributed to detection bias, mainly due to the endoscopies that accompany H. pylori eradication treatment. Because the study’s control group was population-based, exposure to endoscopy should be considerably biased between the eradicated group and the control group. Although we are not familiar with general clinical practice in Sweden in detail, endoscopies may have been performed almost exclusively in the eradicated group, in contrast to the practice in Japan, where nationwide gastric cancer screening is prevalent.2
Endoscopic screening for cancer usually increases the incidence rates due to early detection in younger patients following screening participation, but for the next several years thereafter, the rates drop considerably, reflecting the effect of screening. In colorectal cancers, the cumulative incidence of cancer in the intervention group becomes lower than that of the control group for 5 years after only one flexible sigmoidoscopy examination in the former.3
The reported SIR of 10.74 for the early period (1–3 years) seems too high. The reported prevalence of H. pylori infection in Sweden is around 15%.1 If the risk of gastric adenocarcinoma among patients with H. pylori infection was similar between individuals who received H. pylori eradication therapy and those who did not, the SIR should have been less than 100/15=6.67. (Assuming a risk ratio ‘r’ for gastric cancer due to H. pylori infection, the SIR will be (100×r)/(15×r+85×1)=100/(15+85/r)<100/15.)
Doorakkers et al1 describe a similar trend for cardia adenocarcinoma, which has no association, or even an inverse association, with H. pylori infection.4 We suspect that this result also reflects a screening effect. Verification of our hypothesis requires a similar study in patients with other neoplasms of the upper gastrointestinal tract, such as oesophageal cancer.
More robust outcomes, such as mortality, may better show the effect of H. pylori eradication. Alternatively, determination of the incidence of gastric cancer after a much longer time may also reveal the true preventive effect of eradication treatment. Nonetheless, in either case, separating the effect of H. pylori eradication from that of endoscopy, even after a single screening, may be difficult if the effect of endoscopy is significant and maintained for a long duration.5
Patient consent for publication Not required.
Contributors The two authors have written, revised and approved the letter.
Competing interests None declared.
Provenance and peer review Not commissioned; internally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.