Virulence factors or ancestral origin of Helicobacter pylori: which is a better predictor of gastric cancer risk?

We have read the recently published article by de Sablet et al with great interest. The article is well structured and has important findings about the ancestral origin of Helicobacter pylori, which can be used as a predictor of gastric cancer risk.¹ de Sablet et al performed multi-locus sequence typing (MLST) to determine phylogeographic variation, which was significantly associated with the different histopathological scores and the prevalence of gastric cancer in the specific regions that they studied. However, we think that the study by de Sablet et al has several limitations.

First, there was insufficient data about other virulence factors of H pylori. We examined the association between virulence factors, including cag pathogenicity island, vacA, babA, iceA and OipA, of H pylori strains isolated from Columbian subjects and clinical outcomes.² We found that although OipA and cag pathogenicity island are linked with each other, only OipA was an independent risk factor for duodenal ulcer. OipA tended to be associated with gastric cancer. Thus, in the study by de Sablet et al, OipA in cagA-positive strains may contribute to phylogeographic variation determined by MLST analysis. Furthermore, the difference in histopathological scores may be because of the different status of OipA. Thus, it is important to determine the correlation between other virulence factors and the phylogenetic tree by performing MLST.

Selection bias is another limitation of the study by de Sablet et al because they focused only on the 64 patients with cagA-positive/vacA s1m1 strains but not on the patients with cagA-negative strains. Subjects infected with hpEurope strains of H pylori showed higher histopathological scores than those infected with hpAfrica1 strains. However, all cagA-negative strains were considered to belong to hpEurope. Subjects infected with cagA-negative strains had very low histopathological scores. hpEurope strains without the presence of cagA can be less virulent. Thus, the presence of cagA rather than the phylogeographic origin is a better predictive factor of gastric cancer.

Our preliminary examination included determination of the phylogeographic origin of H pylori strains from Okinawa, Japan. Most strains from Okinawa were divided into three clusters belonging to hpEastAsia. The phylogeographic origin was significantly associated with clinical outcomes; however, it depended on the status of cagA (East-Asian type, Western type and cagA negative) and vacA (m1 or m2). Gastric cancer was more prevalent in the cluster containing most East Asian-type cagA strains. The results of MLST showed that the prevalence of gastric cancer in each group did not differ when only Western-type cagA strains or cagA-negative strains were used. This also suggests that cagA is a better predictive factor than the phylogeographic origin. Further studies are required to confirm the relationship between the ancestral origin of H pylori and clinical outcomes.