Sirs, We were very interested in the paper by Sheu et al.1 about esomeprazole in triple therapy and the efficacy of Helicobacter pylori eradication relating to CYP2C19 genotypes. Their study design is prospectively randomized for treatment groups, using sophisticated exclusion criteria. Genotypic distribution of CYP2C19 for the general population in Taiwan and subjects selected by these exclusion criteria may not be compatible.
The Hardy–Weinberg principle states that single-locus genotypic frequencies after one generation of random mating can be represented by a binomial function of allelic frequency.2 In a population with continued random mating and without factors that change allelic frequency (selection, genetic drift, gene flow and mutation), the Hardy–Weinberg genotypic proportions will not change over time.2 For a system of two codominant alleles with dominant, heterozygous, or recessive types like the CYP2C19, chi-square could be used to test their genotypic proportions and should show no statistically significant differences.2 We found that there are significant differences between observed and expected proportions in the CYP2C19 genotypes, reflecting the possibility of selection bias of their study, which means the generalization of their results is questionable (Table 1). The proportion of their genotypic distribution is also different from many other studies3, 4 and our data5 (Table 1).
|Sheu et al.1||Chen et al.5|
|Homo-EM (w/w, p2)||91||76||157||156|
|Hetero-EM (w/m, 2pq)||65||95||170||172|
|PM (m/m, q2)||44||29||49||48|
The results of Sheu et al.1 showed that the CYP2C19 genotypes can influence the therapeutic effect of different proton-pump inhibitors in the anti-Helicobacter treatment. However, the extensive metabolizer genotypes must be known first for the drug selection. The genotypic distribution of their study cannot represent the general population of Taiwan.