Helicobacter pylori and Antimicrobial Susceptibility in Children

The role of Helicobacter pylori (H. pylori) in the colonization of the stomach in children with chronic gastritis, peptic ulcer, and possibly gastric carcinoma is well documented (1) and eradication of the bacterium has a great effect on prevention of peptic ulcer relapses in children (2). The recent consensus statements on H. pylori infection in children (1,3,4) indicate that upper gastrointestinal endoscopy with biopsies is the preferred method of investigation for children with upper digestive symptoms suggestive of organic disease. Those statements (1,3,4) also propose a treatment and follow-up strategy for those patients who remain infected after a first attempt at eradication (i.e., a second endoscopy with culture and resistance testing to adapt treatment to anti-microbial susceptibility). In contrast, no proposal has been made recommending primary culture and resistance testing before the first treatment. In this issue of the journal, Crone et al. address the prevalence of primary H. pylori-resistant strains in 117 children and analyzed their evolution with time during the period 1997 to 2000. They show no resistance to amoxicillin, 5% to 25% metronidazole-resistant strains and clarithromycin-resistant strains increase with time from 14.3% to 27.6%. Resistance now involves antibiotics previously devoid of any resistance issues for that bacterium, such as amoxicillin. In our experience with children, we have not seen resistance of H. pylori to amoxicillin before first treatment (5,6); nor is it described in Crone's study or all across Europe, in Spain (7), Germany (8), Portugal (9), and Belgium (10). However, H. pylori-resistant strains resistant to amoxicillin were observed in 15.7% in Mexico (11) and 4.6% in the United States (12,13). Resistance of H. pylori to metronidazole before treatment was as high as 43% in children in our experience (6); other European studies show comparable resistance rates (32% in Germany (8), 25% in Crone's study, 23% in Spain (14), 19% in Portugal (9), and 18% in Belgium (10)). Metronidazole is largely used in the treatment of different infections, from parasites to anaerobes. Iterative treatments for parasitic diseases especially in children originating from Africa may therefore be incriminated in the increased resistance observed in this population (6) with a contribution of intrafamilial transmission of the infection (15). The activity of metronidazole in treating H. pylori is dependent on reduction of its nitro moiety to highly reactive compounds that cause DNA strand breakage. It was recently demonstrated that inactivation of the rdxA gene, which encodes an oxygen-insensitive NADPH nitroreductase, is associated with the development of resistance to metronidazole by H. pylori (16). Extensive sequencing in a series of metronidazole-sensitive and -resistant isolates using an H. pylori mouse model showed that the sequence of the rdxA gene of the metronidazole-resistant isolates contained between one and three frameshift or missense mutations. This suggests that while the development of metronidazole resistance in H. pylori is frequently associated with mutational inactivation of the rdxA gene, other mechanisms of resistance are likely to exist in this bacterium (16). Recent evidence has suggested that inactivation of frxA (NADPH flavin oxidoreductase), fdxB (ferredoxin-like protein), and possibly other reductase-encoding genes may also contribute to the resistant phenotype (17). For clarithromycin, a macrolide that partially carries crossover resistance to other antibiotics belonging to the same class, antibiotic resistance was first reported in 1996 (18). Clarithromycin resistance is due to point mutations in only two sites (A2142G) in the 23S ribosomal RNA sequence, that can be easily detected by molecular methods (16) or by the newly rapid and accurate determination of genotypic clarithromycin H. pylori-resistant strains by fluorescent in situ hybridization (19). A resistance of H. pylori to clarithromycin before treatment was detected in more than 15% of cases in the different pediatric European studies (16.6% in Belgium (10), 21% in France (6), 21.13% in Spain (14), 22% in Germany (8), 23.5% in Poland (20), and 44.8% in Portugal (9)). A similar rate was observed in children from Mexico (21.6% (11)) and a higher one (41%) in United States (13). The resistance rate seems lower in children younger than 10 years (16% to 19% as opposed to 9% in children older than 10) (21). The high level of clarithromycin resistance among pediatric H. pylori strains compared with those isolated in adults suggests the potential noxious role of macrolide overuse in children, especially in Europe. Crone's data and other pediatric studies from Belgium (10), Spain (14), and Mexico (11) indicate that clarithromycin resistant strains increase significantly with time. What will be the final outcome of clarithromycin resistance, at the population or individual level? At the population level, it is likely that increased and widespread use in children may in the long term worsen the figures depicted in this study. In Slovenia, the outpatient consumption of macrolides increased from 1.89 to 3.84 daily doses (DDD)/1000 inhabitants/d from 1994 to 1999; this increase was paralleled by a steady increase in macrolide resistance in S. pyogenes and upper respiratory S. pneumoniae isolates (22). In a Japanese study (23), the proportion with clarithromycin resistance significantly increased from 7% in 1997–1998 to 15.2% (P = 0.003) in 1999–2000 and during the same period the metronidazole resistance rate also increased from 6.6% in 1997–1998 to 12% in 1999–2000 (P = 0.02). The prevalence of clarithromycin and metronidazole was related to the annual consumption of these antimicrobial agents (23). The converse situation is also true. In Finland, after nationwide reductions in the use of macrolide antibiotics for outpatient therapy, there was a significant decline in the frequency of erythromycin resistance among group A streptococci isolated from throat swabs and pus samples (24). At the individual level, as elegantly reviewed by M. J. Blaser (2002), in the absence of antibiotic treatment, colonization of the stomach by H. pylori can last for decades and in the form of competing microbes. Strains of H. pylori are diverse and in a given host, colonizing strains experience with time genetic variation, resulting from point mutations or recombination, across loci located within a genome (intragenomic) or between differing organisms (intergenomic) (25). Comparing the genomic sequence of the bacteria in the same host many years later without antibiotic treatment revealed that the later isolates were, albeit closely related, identical neither to the earlier isolate nor each other with individual and groups of genes missing from the new isolates. On the other hand, H. pylori isolates may develop resistance by point mutation under antibiotherapy pressure such as with clarithromycin. When clarithromycin is withdrawn, the resistant strains compete for years with the sister cells that have spontaneously mutated back to the wild, susceptible type. Bjorkholm et al. (26) found that in vitro, wild-type bacteria outcompeted their resistant sisters in most cases, with notable exceptions where the resistant strain outcompeted, perhaps owing to compensatory mutations in genetic loci functionally linked to the 23S ribosomal RNA resistance phenotype. A 23S ribosomal RNA mutation and the compensatory mutation in the other locus may combine to produce a phenotype that, in certain niches, successfully competes with wild-type cells and with organisms that, owing to reversion, have a mutation in only one of the two (or more) loci (25). Authors now suggest that eradication of H. pylori infection should take place only after anti-microbial testing. The treatment of H. pylori infection, in children (6,27) and in adults (28,29), is influenced by resistance to the antibiotics used, which turn out to be the main risk factor for failure. A prospective study performed with 23 H. pylori-infected children treated with triple therapy including metronidazole, indicated an eradication rate significantly dependent on the susceptibility or the resistance of the infecting organism to metronidazole, 83% vs. 17%, respectively (27). We recently observed a failure to eradicate H. pylori infection in children in up to 30% of cases after a one-week triple therapy using omeprazole, amoxicillin, and clarithromycin (30). Furthermore, when the outcome of H. pylori infection was analyzed in 61 children treated with a triple therapy including clarithromycin, bacterial eradication was obtained in all children with clarithromycin-susceptible strains vs none of the children with clarithromycin-resistant ones (P = 0.0001) (31). Meta-analyses have established that resistance to either the 5-nitoimidazole or macrolide component of the therapeutic regimen is an important predictive factor of eradication failure and is less pronounced for metronidazole than clarithromycin (29). One important point is that antibiotic treatment of H. pylori involves drugs largely in use for other kinds of infections. Thus, H. pylori organisms have an increased likelihood of coming in contact with the drug unintentionally when the drug is given for a respiratory illness for example. For that reason, analysis of the resistance phenomenon must take into account both geographical variations and evolution with time, especially according to local therapeutic trends and official guidelines (21,32). Although testing H. pylori isolates for antimicrobial susceptibility seems appropriate before choosing a treatment and clarithromycin avoided in case of resistance, it must be stressed that susceptibility testing is not currently considered a regular prerequisite for successful eradication of H. pylori. This might change since resistant strains continue to rise (29). As a matter of fact, a clear consensus regarding what defines resistance is also needed before it will be adapted to predict accurately treatment failure. In summary, the high level of pre-treatment resistance to clarithromycin of the H. pylori organisms isolated from children highlights the need to determine the antimicrobial susceptibility of H. pylori, when the technique is available, before the first treatment course. Worldwide, theresistance of H. pylori to clarithromycin is rising significantly with time. Finally, resistance rates in children must be monitored to provide guidelines for therapeutic recommendations.