<i>Helicobacter pylori</i> infection and gastric cancer: Facing the enigmas (part II). Reply to Tokudome <i>et al</i>.

Tokudome et al. commented on our ecologic study published in the International Journal of Cancer1 that it is premature to label as an “enigma” the lack of association between Helicobacter pylori infection and gastric cancer on an area level when aggregates of African and Asian countries are considered. Long ago, others were responsible for coining the “African enigma” and “Asian enigma”, and a PubMed search (using the words African enigma or Asian enigma and Helicobacter pylori) shows that, respectively, Holcombe in 19922 and Miwa et al. in 20023 first used these terms. Our paper does not question the association between H. pylori infection and gastric cancer, and obviously this was not one of our findings. Indeed, we assumed the association as a starting point and concluded that interaction between H. pylori and cigarette consumption might further explain the international variation in gastric cancer, contributing to our understanding of the apparent enigma. Thus, our issue was to explain geographic variability and conflicting associations, not to propose a cause for stomach cancer. The letter calls attention to some important points. Data quality is always a crucial issue, and we discussed the strengths and weaknesses of data sources. Age-adjusted H. pylori prevalence would be a better option if information for standardization was available for every study, and a larger set of countries would increase power, probably without changing the main findings. As shown in Table I, the age-adjusted prevalence for a subset of reports providing enough data to standardize rates are similar to those used in our ecologic study (Pearson correlation coefficient = 0.97). Information on chronic atrophic gastritis could be considered, but progression of chronic H. pylori gastritis appears to be similar in Africa, Europe and South America; thus, additional factors may modulate the progression to cancer.4, 5 Our report discussed the large potential for uncontrolled confounding in the area level association between H. pylori and gastric cancer, and we acknowledge that our regression models, although they explained nearly 40% of the variability, missed important determinants of gastric cancer incidence. We relied on routine information and results from published surveys, and there is no way to overcome the lack of data on such variables for most countries. Salt intake, however, measured as 24 hr urine sodium excretion, correlates with H. pylori prevalence in many regions but not in Africa, where salt excretion is generally low and infection frequent, as shown in Figure 1. Urinary sodium excretion vs. adult H. pylori prevalence, by region. Data sources: urinary sodium excretion [Argentina,6 Canada,6 Colombia,6 Mexico,6 United States,6 China,6 India (average Ladakh7 and New Delhi7), Japan,6 Republic of Korea,6 Belgium,6 Denmark,6 Finland,6 Germany (average Democratic Republic6 and Federal Republic6), Hungary,6 Iceland,6 Italy,6 Poland,6 Spain,6 England and Wales,6 Kenya,7 Nigeria6] and H. pylori [Kenya (dyspeptic patients),8 other countries1]. It is widely accepted that H. pylori isolates from high-risk stomach cancer populations may differ genetically from those in low-risk areas. Unfortunately, the presence of cagA, the only pathogenicity marker available for a relatively large number of countries, may be of limited value to explain the “enigmas” as there is also a high prevalence of CagA-positive strains in noncancer patients and healthy individuals in countries with a low frequency of stomach cancer and high H. pylori prevalence.9, 10, 11, 12 Host genetic susceptibility associated with gastric cancer risk likely contributes to the explanation of the worldwide variation in its frequency, but the available evidence on these topics is limited to a small number of countries, mostly Asian.13 Finally, variability in diagnostic test performance is an additional source of error, but there is no reason to imagine that African or Asian countries present high prevalence rates of infection due to poorly performing tests. Thus, we thank Tokudome et al. for pointing out that there is no obscurity in the so-called enigmas and proposing new individual-level research to confirm our group analysis results. Yours sincerely, Nuno Lunet, Henrique Barros

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.