Antibacterial resistance and the success of tailored triple therapy in Helicobacter pylori strains isolated from Slovenian children

Abstract Background Primary Helicobacter pylori (H. pylori) infection occurs predominantly in childhood. Antimicrobial resistance is the leading cause for H. pylori eradication failure. The aims of this study were (i) to establish for the first time the antimicrobial resistance of H. pylori strains in infected Slovenian children not previously treated for H. pylori infection and (ii) to evaluate the effectiveness of tailored triple therapy, assuming that eradication rate with tailored triple therapy will be >90%. Methods Data on all treatment‐naive children 1‐18 years old and treated for H. pylori infection according to susceptibility testing were retrospectively analyzed. All relevant clinical information and demographical information were retrospectively collected from the hospital information systems and/or patients’ medical documentation. Results The inclusion criteria were met by 107 children (64.5% girls) with a median age of 12.0 years (range 2.0‐17.6 years). Primary antimicrobial resistance rates of H. pylori were 1.0% to amoxicillin (AMO), 23.4% to clarithromycin (CLA), 20.2% to metronidazole (MET), 2.8% to levofloxacin (LEV), and 0.0% to tetracycline (TET). Dual resistances were detected to CLA and MET in 11.5% (n=12) of strains, to CLA and LEV in 2.8% (n=3), and to MET and LEV in 2.9% (n=3). Results of treatment success were available for 71 patients (66.2% girls). Eradication of H. pylori was evaluated using the 13C‐urea breath test, monoclonal stool antigen test or in some cases with repeated upper GI endoscopy with histology and cultivation/molecular tests. Eradication was achieved in 61 of 71 (85.9%) patients. Conclusions The primary resistance rates of H. pylori to CLA and MET in Slovenia are high. Our data strongly support the fact that in countries with high prevalence of resistant H. pylori strains susceptibility testing and tailored therapy is essential.


| INTRODUCTION
Infection with H. pylori is still the most common bacterial infection in humans, despite the significant worldwide decrease in its incidence and prevalence recorded in recent years. 1 Chronic inflammation of the gastric mucosa can cause peptic ulcers, gastric erosions, and mucosaassociated lymphoid tissue lymphoma in children. In addition, extra gastric complications, such as iron-deficiency anemia, may be induced by H. pylori infection. 2 The first-line treatments currently available for children infected with H. pylori are various combinations of protonpump inhibitors (PPI) with antibiotics including AMO, macrolides, and/ or imidazoles. 3,4 In the last decade, the eradication rates using these schemes have declined and have led to recommendations for higher dosages and longer duration of therapy or alternative regimens, including three instead of two different antibiotics, either sequentially or concomitantly. 4,5 The low success rates are essentially due to the increasing resistance rates to macrolides and to a lesser extent imidazoles. Koletzko et al. 6 detected very high primary resistance rates to metronidazole (MET) and clarithromycin (CLA) in Europe (23% and 20%, respectively), and resistance after failed eradication therapy was even higher (35% and 42%, respectively). Children born to mothers from developing countries have extremely high primary resistance rates to MET presumably due to the frequent use of this drug for parasitic infections in Africa and Asia. Higher primary CLA resistance rates in pediatric patients compared to adults are thought to be a consequence of frequent use of macrolides for treatment of respiratory infections. Therefore, in order to improve eradication rates in children infected with H. pylori, the latest ESPGHAN and NASPGHAN guidelines for developed countries suggest eradication therapy according to the results of antimicrobial susceptibility testing. Namely, antibiotic susceptibility testing for CLA is recommended before initiation of CLA-based triple therapy in areas/populations with a known high resistance rate (>20%) of H. pylori to CLA or where antibiotic resistance rates are unknown. 3 The aims of this study were (i) to establish for the first time the antimicrobial resistance of H. pylori strains among Slovenian children not previously treated for H. pylori infection and (ii) to evaluate the effectiveness of tailored triple therapy, assuming that eradication rate with tailored triple therapy will be >90%. 7

| Study design and patients
We retrospectively analyzed data for all treatment-naive children guidelines, which recommend that initial diagnosis of H. pylori infection is based on either positive histopathology and a positive RUT or a positive culture. 3 The success of eradication was evaluated with the 13C-urea breath test (UBT) or monoclonal stool antigen test at least 4 weeks after the treatment. Previous eradication therapy, the use of PPI, anti-secretory drugs, or antimicrobial agents 1 month prior to the upper endoscopy, or an eradication protocol not in accordance with antimicrobial susceptibility results was exclusion criteria.
The study was approved by the Republic of Slovenia National Medical Ethics Committee, no. 153/06/14.

| Biopsy sampling and antimicrobial susceptibility testing
Biopsies were taken during upper endoscopy from the duodenum, gastric antrum and corpus, and distal esophagus for histology; additional biopsies were taken from the antrum for RUT and cultivation.
The antral biopsies were collected and transported to the Institute of Microbiology and Immunology, Faculty of Medicine, University of Ljubljana. Biopsy specimens were homogenized and plated on three agar media: Pylori agar (bioMérieux, Marcy l'Etoile, France) and two in-house agar media enriched with 9% horse serum and 9% human blood and incubated for 3-9 days at 37°C in a microaerophilic atmosphere. Identification of H. pylori was confirmed using typical colony morphology, urease, catalase, and oxidase tests. Culture-positive biopsy specimens were phenotypically tested for susceptibility to AMO, CLA, MET, levofloxacin (LEV), and tetracycline (TET) using the gradient-diffusion method, ISO Sensitest agar supplemented with defibrinated horse blood, and EUCAST breakpoints tables version 4.0 for interpretation criteria based on epidemiological cutoff values, which distinguish wild-type isolates from those with reduced susceptibility.
Strains were resistant to antimicrobial agent when minimal inhibitory concentrations were >0.125 mg/L for AMO, >0.5 mg/L for CLA, >8 mg/L for MET, and >1 mg/L for LEV and TET. All culture-negative specimens were further tested for susceptibility to CLA using two molecular assays: the in-house real-time PCR targeting well-described 23S rRNA mutations and the Genotype HelicoDR (Hain Lifescience GmbH, Nehren, Germany) commercial assay with the same target sites.

| Follow-up to assess the success of guided triple therapy
The eradication rate was evaluated using the UBT, a stool antigen test, or in some cases repeated upper GI endoscopy with histology and cultivation/molecular tests 8 weeks after the tailored therapy. The cutoff value of the UBT used (Helicobacter test INFAI, INFAI UK Ltd., York, UK) was 4‰. According to the manufacturer's instructions, the results of the stool antigen test were defined as negative, positive, or indeterminate. Because the results of the UBT and stool antigen test were often reported only to the child's primary pediatrician and not to the pediatric gastroenterologist that prescribed the eradication therapy, we inquired about those results in writing or with a phone call.

| Statistical analysis
The results were collected into an electronic database (Excel, Microsoft, USA). Statistical analysis was performed using IBM SPSS ver. 21 (IBM Corporation, Armonk, NY, USA). Categorical variables were summarized as frequencies and percentages; the median and range were calculated for the patient's age. Univariate logistic regression was performed with H. pylori eradication as a dependent variable and sex, age groups (1-11 or 12-18 years old), duration of antimicrobial therapy, and therapy with CLA or MET as covariates. Odds ratios (ORs) and 95% confidence intervals were calculated. A P-value <.05 indicated significant differences.
Nodular gastritis was the predominant endoscopic finding present in 98.1% (n=105) of patients, erosive gastritis was present in 14.0% (n=15), 3.7% (n=3) of patients had a gastric ulcer, and 0.9% (n=1) had a duodenal ulcer. The population characteristics and endoscopic data are summarized in Table 1.

| Antimicrobial resistance of Helicobacter pylori
Altogether, 107 strains from treatment-naive children were tested for antimicrobial susceptibility, of which 104 were culture-positive and were consequently tested with phenotypic methods. Three strains were culture-negative. For these three strains, susceptibility testing was performed with genotypic methods, and only results for CLA and LEV were available. Primary resistance rates were 23.4% (n=25) to CLA, 20.2% (n=21) to MET, 2.8% (n=3) to LEV, 1.0% (n=1) to AMO, and 0.0% (n=0) to TET. Dual resistance was detected to CLA and MET in 11.5% (n=12) of strains, to CLA and LEV in 2.8% (n=3), and to MET and LEV in 2.9% (n=3). Triple resistance to CLA, MET, and LEV was detected in 2.9% (n=3) of strains, and quadruple resistance to CLA, MET, LEV, and AMO in 1.0% (n=1). Approximately 68.3% (n=71) of H. pylori strains were sensitive to all antimicrobial agents tested. The data are shown in Table 2.

| Antimicrobial therapy eradication rate
Inclusion criteria for evaluation of antimicrobial therapy were met in 66.4% (n=71) children (per-protocol population). Among the excluded patients, 15.0% (n=16) did not respond to the questionnaire, 6.5% (n=7) did not perform evaluation of therapy with the UBT or stool antigen test, 6.5% (n=7) did not receive therapy according to antimicrobial susceptibility results, and 5.6% (n=6) were lost to follow-up for other reasons. (n=61) of children. Using univariate logistic regression, there was no statistical significant difference in eradication with regard to sex, age group, treatment duration, and antimicrobial therapy using CLA or MET (P>.05). The eradication rates are shown in Table 3.
None of the patients stopped their therapy due to adverse effects. This is the first study that evaluated the success rate of H. pylori eradication in Slovenian children. Our study detected high primary resistance rates for both CLA and MET and substantial dual or multidrug resistance rate ( Table 2). The eradication rate with tailored therapy was 85.9% and did not achieve the target rate of >90%. However, the latter target was set based on dual resistance rate of <15% 7 to which our dual resistance rate came fairly close at 11.5%. Therefore, the below expected eradication rate might have been expected.

| DISCUSSION
The first aim of this multicenter study was to establish the pri- to CLA in their study was 12% and they attributed this to relatively low macrolide consumption in Bulgaria. 10 The lowest primary CLA resistance (1%-5%) was reported in the Netherlands. 11,12 Janssen et al. 11 believe that these differences in primary resistance to CLA in the Netherlands compared to other European countries may be related to the extremely low use of antibiotics. We believe that the reason  Several studies on the antimicrobial resistance of H. pylori among European children have shown primary resistance to MET at 10%-35%, to AMO at 0%-1%, and to LEV at 0.2%-0.6%, 6,9,10,14,15 whereas resistance to TET has not generally been reported in Europe. 14 Our results for AMO, MET, and TET are in line with these results, whereas resistance to LEV (2.8%) was slightly higher, but we do not have a good explanation for the difference in results compared to neighboring countries.
Dual resistance to CLA and MET in treatment-naive patients was reported in up to 17% of children in recent reports. 6,9,15,16 In our study, an 11.5% (n=12) dual resistance rate was detected, which is in line with the published data.
When evaluating the eradication rate of tailored triple therapy according to susceptibility testing among children in Slovenia, the eradication rate did not achieve 90% as a goal. 7 The overall eradication rate in this study was 85.9%. It was lower in comparison with previously published results with a >90% and even >95% eradication rate. [17][18][19] Various factors can be involved in the lower success of treatment in our study, such as antibiotic resistance, mixed infections, and the duration of therapy. Furthermore, our study was retrospective; therefore, we were unable to ensure that all the patients took their eradication therapy as prescribed.
It is well known that in vitro antimicrobial susceptibility does not always correspond to successful eradication in vivo. 20 23 The study published by our group showed that 12% of Slovenian children were infected with more than one strain, determined by genetic testing. 24 Therefore, a better eradication rate was probably achievable in our study by obtaining multiple specimens from different parts of the stomach for cultivation.
The success rate could also be influenced by duration of treatment.
Data regarding the optimal duration of therapy for H. pylori eradication are conflicting. A meta-analysis performed by Ford et al. tried to define the optimum duration of PPI-based triple therapy among adults.
Studies comparing 7-day with 14-day treatment with an identical drug combination showed that eradication rates were higher with longer duration of therapy. They concluded that treatment duration should be tailored according to the clinical situation. 25 The authors of the Cochrane review, which included 75 studies, concluded that increasing the duration of PPI-based triple therapy increases eradication rates among adults. 26 The results in adults were not confirmed by the systematic review of H. pylori eradication treatment schedules in children published by Oderda et al. 27 She concluded that longer courses of PPI-based triple therapies are not more effective than shorter ones.
Our study, limited by a small sample size, did not show any statistical significant difference in eradication rate when comparing 7-and 10vs 14-day treatment schemes (P=.62).
Our study has several limitations. It was a retrospective study with all of the inherent limitations of this design. It was difficult to obtain the eradication data for all of the patients included. Namely, almost onethird of patients were lost to follow-up. In addition, although it was a multicenter study, it does not completely represent the Slovenian pediatric population infected with H. pylori because we were not able to obtain data from all Slovenian hospitals. Furthermore, the overall eradication rate was based on different durations of therapy, different antibiotic regimens and varied follow-up tests to assess H. pylori eradication.
In conclusion, this is the first study that reports primary resistance and eradication success of tailored triple therapy for H. pylori infection among the Slovenian pediatric population. High CLA (23.4%) and MET (20.2%) primary resistances were detected. Our data strongly support the fact that in countries with high prevalence of resistant H. pylori strains susceptibility testing prior to eradication treatment is essential. Although triple therapy tailored to antimicrobial susceptibility was highly successful in our study, an eradication rate greater than 90% was not achieved.