Stomach Bug in Children Shows Alarming Antibiotic Resistance, Study Finds

A sobering picture of antibiotic resistance in childhood stomach infections has emerged from a new study of Chinese children, where more than eight in ten bacterial isolates resisted two of the most commonly used antibiotics and nearly three quarters resisted at least two drugs at once. The research, published in Pediatric Research by a team at Shengjing Hospital of China Medical University, examined how well laboratory tests and genetic screening could predict which treatments would actually cure children infected with Helicobacter pylori, the bacterium that colonizes the stomach lining and can cause gastritis, ulcers, and in some cases longer-term digestive disease.

The study took a retrospective look at children who had undergone clinically indicated endoscopy between October 2022 and October 2024. In these procedures, physicians obtained gastric biopsy samples that could be tested in two complementary ways. First, the team used polymerase chain reaction, or PCR, followed by Sanger sequencing, to search the bacterial genetic material for known resistance-associated variants in key genes. Second, they attempted to grow live bacteria from the biopsies in culture, and when that succeeded, they subjected the isolates to antimicrobial susceptibility testing, the traditional method of exposing bacteria to antibiotics and measuring whether the drugs inhibit growth.

Of the 145 children enrolled in the analysis, 128, or 88.3 percent, tested positive for H. pylori by PCR, while 92 yielded positive cultures. The phenotypic resistance figures that emerged from susceptibility testing were striking. Clarithromycin, a macrolide antibiotic that forms the backbone of many standard eradication regimens, failed to inhibit 82.6 percent of isolates. Metronidazole, another frequently used agent, was ineffective against 81.5 percent. Levofloxacin, a fluoroquinolone often held in reserve for second-line therapy, showed resistance in 29.3 percent of isolates. Most concerning of all, 72.8 percent of the cultured isolates were resistant to at least two of these antibiotics, a level of multidrug resistance that sharply narrows the menu of empirical options available to clinicians who must choose treatment before any laboratory results return.

The molecular side of the study focused on two genes with well-established roles in resistance. Mutations in the 23S ribosomal RNA gene are the principal drivers of clarithromycin resistance, altering the binding site on the bacterial ribosome so that the drug can no longer interfere with protein synthesis. Variants in gyrA, which encodes a subunit of the enzyme DNA gyrase, underpin levofloxacin resistance by changing the drug’s target in the bacterial replication machinery. When the researchers compared PCR-detected mutations in these genes against the phenotypic test results, the agreement was remarkably strong. For clarithromycin nonsusceptibility, molecular prediction achieved 98.8 percent sensitivity, 90.9 percent specificity, and 97.8 percent accuracy. For levofloxacin, the corresponding figures were 90.9 percent sensitivity, 97.7 percent specificity, and 95.5 percent accuracy.

Those performance statistics matter because they speak to a long-standing practical dilemma in pediatric gastroenterology. Culture-based susceptibility testing, while definitive, is slow, technically demanding, and not always successful; in this cohort, cultures were positive for only 92 of the 145 children despite nearly universal PCR positivity. Molecular testing, by contrast, can be completed quickly from biopsy material even when the organism cannot be grown, and it can detect heterogeneous mixtures of resistant and susceptible organisms within a single sample. The new data suggest that for clarithromycin and levofloxacin, the genetic readout is a close and reliable proxy for what the phenotype would show, making rapid PCR-based screening a realistic complement to culture in routine care.

The picture was less clear-cut for other antibiotics. The study’s authors note that molecular findings for agents beyond clarithromycin and levofloxacin were less readily interpretable, which reflects the messier biology of resistance to drugs such as metronidazole. Resistance to nitroimidazoles like metronidazole involves multiple genes, including those encoding enzymes that activate the drug inside the bacterial cell, and no single mutation reliably predicts failure. That biological complexity explains why the team did not claim equivalent molecular performance across the board, and why culture remains indispensable for antibiotics whose resistance mechanisms are polygenic or incompletely mapped.

Treatment outcomes provided the clinical payoff of the analysis. Among 97 children who received eradication therapy and completed follow-up, 86, or 88.7 percent, were cured as confirmed by the carbon-13 urea breath test, a noninvasive test in which the child drinks a labeled urea solution and breath samples reveal whether urease-producing bacteria remain in the stomach. Critically, the study linked resistance data to prescribing behavior: of the 84 culture-positive children who were treated, 80 received regimens that were concordant with the phenotypic susceptibility results. That alignment between laboratory findings and drug selection is precisely what the susceptibility-guided approach is meant to achieve, and the high eradication rate suggests it worked in practice.

The findings land in a field already wrestling with troubling regional and global trends. Systematic reviews cited by the authors document rising primary antibiotic resistance in H. pylori across the Asia-Pacific region and among children worldwide, with clarithromycin resistance in East Asian pediatric populations climbing to levels that undermine standard triple therapy. Updated joint guidelines from the European and North American pediatric gastroenterology societies, ESPGHAN and NASPGHAN, released in 2023, emphasize the importance of antimicrobial susceptibility testing where possible, and a 2025 study found that susceptibility-guided treatment outperformed empirical therapy in children. The new Chinese data add granular, genotype-to-phenotype-to-outcome evidence supporting that direction of travel.

For clinicians, the practical message is twofold. Where culture and susceptibility testing are available, they should inform regimen selection, particularly in settings where resistance prevalence is high enough that empirical choices are likely to miss. Where culture fails or delays are unacceptable, rapid molecular testing of 23S rRNA and gyrA can reliably flag clarithromycin and levofloxacin resistance, allowing physicians to steer away from doomed regimens before the first dose is given. The study also underscores the limits of either approach alone: molecular panels cannot yet substitute for phenotypic testing across all antibiotics, and culture cannot always be obtained. The complementary strategy demonstrated here, in which each method covers the other’s blind spots, achieved an eradication rate near 89 percent in a cohort with extraordinary baseline resistance.

The broader implications extend beyond a single hospital in Shenyang. H. pylori infects a substantial fraction of children globally, and childhood infection is the reservoir from which adult disease, including gastric cancer risk, ultimately springs. Every failed treatment course exposes the patient to unnecessary drugs, drives further resistance selection, and leaves the infection in place. As resistance rates continue to climb, the kind of paired molecular and phenotypic diagnostics validated in this study may shift from a research luxury to a standard of care, particularly in high-prevalence regions where empirical therapy has become a gamble. The authors’ demonstration that genetic screening closely mirrors phenotypic reality for the two most actionable resistance targets offers a concrete, technically grounded path toward smarter, faster, and more successful treatment of one of the world’s most persistent childhood infections.

Subject of Research: Antimicrobial resistance and molecular resistance prediction in pediatric Helicobacter pylori infection

Article Title: Antimicrobial resistance patterns, molecular resistance prediction, and treatment outcomes in pediatric Helicobacter pylori infection

Article References: Chen, Y., Gu, Z., Sun, M., Mao, Z., & Guo, J. (2026). Antimicrobial resistance patterns, molecular resistance prediction, and treatment outcomes in pediatric Helicobacter pylori infection. Pediatric Research. https://doi.org/10.1038/s41390-026-05435-0

Image Credits: AI Generated

DOI: 10.1038/s41390-026-05435-0

Keywords: Helicobacter pylori, antimicrobial resistance, pediatrics, clarithromycin, metronidazole, levofloxacin, PCR, 23S rRNA, gyrA, eradication therapy, susceptibility testing, urea breath test