Clarifications Regarding the 3′ Repeat Region of the cagA Gene in Helicobacter pylori and Clinical Outcome

In a recent issue, Rota et al. (2) reported a study regarding the 3′ repeat region of the cagA gene of Helicobacter pylori in which they classified the cagA 3′ region into three types (A, B/D, and C) according to our previous report (5). They reported no association between 3′ repeat region subtypes and clinical outcome. However, multiple subtypes of the cagA repeat region were associated with gastric ulcer. Unfortunately, the PCR primers they used are located outside of the repeat regions, and strains isolated from East Asian and non-Asian countries are detected by PCR using these primers. As we previously described, the sequences of the second repeat regions in strains from East Asia are completely different from those in strains from non-Asian countries (3, 4). Non-Asian strains possess 102-bp second repeat regions, and East Asian strains possess 162-bp second repeat regions (3–5). Therefore, subtypes A through D cannot be applied to non-Asian strains. In East Asian strains, types A (around 648 bp) and B (around 756 bp) have one second repeat region and types C (around 810 bp) and D (around 756 bp) have two second repeat regions. The expected lengths of the PCR products in East Asian strains are equal to (423 ∼ 432) + (57 × f) + (162 × s), where f is the number of first repeat regions and s is the number of second repeat regions. In contrast, the expected lengths of PCR products in non-Asian strains are equal to (486 ∼ 498) + (57 × f) + (102 × s) bp. We found that typically (>95%) there was only one first repeat region in non-Asian strains (3, 4) such that the expected lengths of the PCR products are (543 ∼ 555) + (102 × s) bp. Thus, 650-, 750-, and 850-bp fragments indicate H. pylori with one, two, and three cagA non-Asian type second repeat regions, respectively, which is in complete agreement with data by Rota et al. confirming that the Brazilian strains have a non-Asian cagA structure. We reported that non-Asian H. pylori strains with three or more second repeat regions were associated with gastric atrophy and intestinal metaplasia (3). Recently, Kidd et al. reported that, compared to the prevalence in patients with gastritis alone, the prevalence of the shortest PCR fragment in the 3′ region of the cagA gene was high in peptic ulcer patients and the prevalence of the longest fragment was high in patients with gastric cancer (1). Rota et al. also noted that the presence of multiple genotypes in the cagA repeat region was associated with gastric ulcer (which is typically associated with pangastritis and some atrophy) but not with duodenal ulcer (which typically has a nonatrophic gastritis) (2). Although neither Rota et al. nor Kidd et al. provided data about the presence of atrophy, both reports are consistent with the hypothesis that cagA second repeat regions are associated with the premalignant condition of gastric atrophy. Finally, the “Japanese methodology” was developed in Houston, Tex. (3).