Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Pooled heterogeneous studies show an association between Helicobacter pylori infection and gastric cancer


ACP J Club. 2000 Mar-Apr;132:74. doi:10.7326/ACPJC-2000-132-2-074

Source Citation

Eslick GD, Lim LL, Byles JE, Xia HH, Talley NJ. Association of Helicobacter pylori infection with gastric carcinoma: a meta-analysis. Am J Gastroenterol. 1999;94:2373-9. [PubMed ID: 10483994]



Is there an association between Helicobacter pylori infection and gastric cancer?

Data sources

Studies were identified by searching MEDLINE, CINAHL, CANCER CD, Biological Abstracts, and Current Contents (1983 to March 1999); scanning reference lists of selected articles; and contacting international experts in epidemiology and gastroenterology.

Study selection

Studies in any language were selected if H. pylori infection was the exposure variable, gastric cancer was the disease of interest, and the publication date was after 1982.

Data extraction

Reviewers were given only the methods section of selected articles, and they assessed the quality of study methods by using a coding manual developed for this purpose (maximum score 13). Data were extracted on study design, country, diagnostic methods, year of publication, sex, age, stage of cancer, number of cases and control patients, and H. pylori status.

Main results

42 studies (8 cohort and 34 case-control studies) met the selection criteria. 7 of 22 studies showing a positive association had high-quality scores (scores 10 to 13) compared with 1 of the 20 negative studies. The pooled results showed an association between the presence of H. pylori infection and an increased risk for gastric cancer (combined odds ratio [OR] 2.04, 95% CI 1.69 to 2.45). Heterogeneity existed among the studies (P < 0.001). In multivariate analyses, age and intestinal-type cancers were independent effect modifiers. Younger participants had a lower risk for gastric cancer than did older participants (combined OR 0.77, CI 0.68 to 0.89). The association between H. pylori infection and gastric cancer was stronger for intestinal cancer than for diffuse cancer (OR 1.14, CI 1.05 to 1.25).


Meta-analysis of heterogeneous studies shows an association between Helicobacter pylori infection and gastric cancer.

Source of funding: Not stated.

For correspondence: Dr. N.J. Talley, Department of Medicine, University of Sydney, Nepean Hospital, Clinical Sciences Building, P.O. Box 63, Penrith, New South Wales 2751, Australia. E-mail


Decreased incidence of gastric cancer in many countries is attributed at least in part to a reduced rate of H. pylori infection. Although H. pylori has been classed as a carcinogen (1), evidence that H. pylori infection is more common than expected in patients with gastric cancer is mixed.

Eslick and colleagues have combined data from 42 studies that were selected using appropriate inclusion criteria. They conclude that gastric cancer is twice as common in persons with evidence of H. pylori infection as in uninfected persons. The country of origin of the data, the type of controls (inpatients or non-inpatients), or the type of diagnostic test did not explain differences.

These results suggest that infection with H. pylori is an important determinant of risk for gastric cancer and that a decrease in gastric cancer incidence may indeed be attributed to decreasing rates of H. pylori infection. However, some oddities in the findings qualify any important clinical messages that might be taken from this review.

First, gastric cancer tends to be of 2 types. One type is proximal, diffuse in histology, more common in men than women, and probably rising in incidence in Western countries (as is esophageal adenocarcinoma) (2). The other type tends to be distal, intestinal in appearance, equally common in men and women, and decreasing in incidence. The suspicion has been that this latter type is associated with H. pylori; but the meta-analysis, which shows a greater, although very modest, increase in risk for intestinal cancer than for diffuse cancer, shows no relation with site.

Second, risks measured as ORs vary widely, even in studies with large numbers of patients. For instance, the OR in Finland was 1.31 (CI 0.99 to 1.74) for 243 case patients and 1408 control patients, whereas in Italy, the OR was 3.66 (CI 2.33 to 5.74) for 307 case patients and 162 control patients. Gastric cancer is a disease of poverty, and 1 possible explanation is that without matching for social class, the true risk for persons with H. pylori compared with those without may not be easily measured.

The third consideration is whether the particular variety of H. pylori matters (3). The answer is “possibly,” but the results cannot tell us whether eradicating established infection will prevent gastric cancer. The risk for esophageal cancer might also increase if eradication makes patients more prone to gastroesophageal reflux. Because esophageal and gastric cancers are about equally common now (4), the equation may be finely balanced. More evidence is needed on this matter.

Michael J. Langman, FMedSci, MD
University of Birmingham
Birmingham, England, UK


1. Infection with Helicobacter pylori. IARC Monogr Eval Carcinog Risks Hum. 1994;61:177-240. [PubMed ID: 7715070]

2. Blot WJ, Devesa SS, Kneller RW, Fraumeni JF Jr. Rising incidence of adenocarcinoma of the esophagus and gastric cardia. JAMA. 1991;265:1287-9. [PubMed ID: 1995976]

3. el-Serag HB, Sonnenberg A. Opposing time trends of peptic ulcer and reflux disease. Gut. 1998;43:327-33. [PubMed ID: 9863476]

4. Devesa SS, Blot WJ, Fraumeni JF Jr. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer. 1998;83:2049-53. [PubMed ID: 9827707]