Antibiotic resistance characteristics and risk factors analysis of Helicobacter pylori strains isolated from patients in Liaoning Province, an area in North China

Background The prevalence of Helicobacter pylori (H. pylori) keeps rising while the eradication rate continues to decline due to the increasing antibiotic resistance. Regional variations of antimicrobial resistance to H. pylori have been recommended by guidelines in recent years. This study aims to investigate the antibiotic resistance rate of H. pylori and its association with infected subjects’ characteristics in Liaoning Province, an area in north China. Methods Gastric tissues from 178 H. pylori positive participants without previous antibiotic use within four weeks were collected for H. pylori culture. Antibiotic susceptibility to furazolidone (AOZ), tetracycline (TC), levofloxacin (LFX), metronidazole (MET), clarithromycin (CLA), and amoxicillin (AMX) were examined with the agar dilution method. Associations between H. pylori resistance and patient characteristics were further analysed. Results No resistance was observed in AOZ or TC. For LFX, MET, CLA, and AMX, the overall resistance rates were 41.10%, 79.14%, 71.78%, and 22.09% respectively. There were significant differences between resistance to CLA and MALToma (P = 0.021), and between resistance to MET and age (P < 0.001). Conclusions The primary resistant rates of LEX, MET, CLA, and AMX were relatively high in Liaoning. Treatment effectiveness improvement could be achieved by prior antimicrobial susceptibility tests before antibiotic prescription.


INTRODUCTION
Helicobacter pylori (H. pylori) is a ubiquitous bacterium which is well resistant to the gastric acid. It was estimated that the global prevalence of H. pylori was approximately 50% (Hooi et al., 2017). Previous studies established that the colonization of H. pylori was directly involved in the occurrence of various gastrointestinal diseases (Kishikawa et al., 2020;Yamaoka, 2018). Emerging evidence has suggested that H. pylori eradication could cure

Selection of patients
The enrolled subjects were all patients at the gastroenterology department of the First Affiliated Hospital of China Medical University, a tertiary hospital in Liaoning Province during March 2021 to August 2022. Patients enrolled came from various cities in Liaoning Province. Every selected patient was H. pylori positive and has not received any antibiotic use in the last four weeks. Subjects positive for H. pylori infection, aged ≥18 years were recruited, with at least one of the following diseases: superficial gastritis (SG), atrophic gastritis (AG), gastric cancer (GC), mucosa associated lymphoid tissue lymphoma (MALToma) and peptic ulcer diseases (PUD). The exclusion criteria included subjects administered any antibiotics in the past 4 weeks.

Acquisition of clinical specimens and Basic information
Gastric mucosa tissues of the subjects were obtained through the endoscopic biopsy process to confirm the infectious status of H. pylori and to perform the antibiotic sensitivity test. Basic information of the subjects was obtained from HIS Information System Technical Support Services, including age, gender, current smoking status, current drinking conditions, BMI, hypertension, diabetes mellitus, endoscopic diagnosis, and H. pylori treatment history. The definition of current smoking and drinking in the present study is that participated subjects engaged in smoking and/or drinking behaviour in accordance with their personal smoking and/or drinking habits during the first 4 weeks before conducting the drug sensitivity test. This project was approved by the Human Ethics Review Committee of the First Affiliated Hospital of China Medical University (2021325). We have received written informed consent from participants.

Determination of H. pylori infection
Urea breath test (UBT) was performed with The Kit For 13 C-Urea Breath Test (Haiderun Pharmaceutical Group Co. LTD., Beijing, China) through intaking 100-mg 13 C-labelled reagent. The WLD600C 13 C Analyser (Haiderun Pharmaceutical Group Co. LTD., Beijing, China) was applied to analyse the breath samples and a positive result was decided following instructions of the manufacturer. Three gastric biopsy specimens of each subject (one from the antrum, one from the angle and the other from the corpus) were placed in a H. pylori transport culture medium (patent number: CN104762235A) (Guo, 2018) and transferred on ice to the H. pylori laboratory at Chain Medical Labs in Changchun, China (91220101MA0Y6MJX1T). Then the samples were fully grounded and inoculated into Columbia blood agar (CBA) plates, supplemented with 5% sheep blood. The colony of H. pylori was observed after being cultured on the plate under microaerobic conditions (5% oxygen, 85% nitrogen, and 10% carbon dioxide) at 37 • C for three to four days. Positive urease, catalase and oxidase tests and Gram-staining were performed to identify colonies resembling H. pylori (Blanchard & Nedrud, 2012). The patient who received positive results from both UBT and culture was identified to be H. pylori-positive.

Statistical analyses
All statistical analyses were performed using the statistical software SPSS (version 18.0; SPSS Inc., Chicago, IL, United States). Categorical data were presented as number and percentage and continuous data were presented as mean ± standard deviation (SD). The following characteristics data from participated subjects were processed using descriptive statistical analysis, including age, gender, current smoking status, current drinking conditions, BMI, hypertension, diabetes mellitus, endoscopic diagnosis, H. pylori treatment history, and antibiotic susceptibility. Chi-square test was applied to compare the difference between groups. Fisher's exact test was applied when over 20% of the expected counts were below 5. P value less than 0.05 in two tails was considered statistically significant.

RESULTS
Among 178 patients determined as H. pylori positive by Urea breath test (UBT), successful bacterial cultures were obtained for 163 (91.57%) patients.

Antibiotic resistance rates in H. pylori isolates
No resistance to AOZ or TC was observed in the population. The overall resistance rates to both MET and CLA remained at about the same high level (79.14% for MET, followed

Relationships between H. pylori antibiotic resistance and patient characteristics
The results of Fisher's exact test suggested that resistance to CLA and endoscopic findings (P = 0.010) significantly differ (Table 3). We further explored the relationship between CLA-resistance situation and every endoscopic finding by chi-square test. A significant difference was found between CLA-resistance and MALToma (P = 0.021; Table 4). Another difference existed between MET-resistance and age (P < 0.001). No other correlation was discovered between antibiotic resistance and patient characteristics including gender, current drinking conditions, BMI, hypertension, diabetes mellitus and H. pylori treatment history (P > 0.05; Table 3).

DISCUSSION
H. pylori antibiotic resistance is a major threat to affect the success of current therapies and regional difference has been an ordinary issue in the antibiotic resistance to H. pylori. This issue could be alleviated by antimicrobial susceptibility testing for H. pylori, but with harsh growth environment requirements, operational safety and economic benefits and risks to be further evaluated, so this testing method almost remains unavailable in clinical practice (Jiao, Wang & Ma, 2021;Jones et al., 2017). Under this circumstance, profiling regional antibiotic resistance patterns and specific characteristics is an extremely important measure for guiding the choice of therapeutic regimens. It is urgent to supervise respective antibiotic resistance patterns in different areas to solve the problem of indiscriminate and excessive use of antibiotics. The resistance rate to antibiotics of H. pylori in China has been increasing with an upward trend during several years. From two studies, primary resistance rates to LFX, MET, CLA and AMX were all much higher than 5 years ago (Chen et al., 2022;Hu, Zhu & Lu, 2017) , as evidence of the difficult eradication today. The data about antimicrobial resistance in Liaoning were the first to be reported and revealed a situation that the overall resistance rates to LFX, MET, CLA, and AMX were higher than those in China (Chen et al., 2022) , especially for CLA (71.78% further higher than 30%), except for TC and AOZ with undetected results. These results were consistent with the conclusion by Liu et al. (2019). Such different prevalence might be correlated well with urbanization level, socioeconomic status, sanitary level and so on (Hooi et al., 2017). We found that both primary and overall antibiotic resistance rates to CLA (70.59% and 71.78%, respectively) were much higher than those in China (34% and 30%, respectively) (Chen et al., 2022) , possibly because the cold weather in north areas could cause colds and flu, and macrolides are widely prescribed for respiratory infections (Megraud et al., 2021). CLA is still widely used in many regions as an antibiotic in the standard triple therapy for H. pylori eradication (Fallone et al., 2016). It was recommended to avoid the usage of CLA if the local resistance to CLA was over 15% (Malfertheiner et al., 2017;Mascellino et al., 2017). Therefore, in the studied region, regimens including CLA was not appropriate for empirical treatment unless the prior antibiotic susceptibility test revealed a sensitive result with CLA. For AMX, all the AMX-resistant patients were multi-resistant. We found the primary AMX-resistant  Notes. Significant differences were found between CLA-resistance and endoscopic findings (P = 0.010), and between MET-resistance and age (P < 0.001).  Notes. Further significant difference was found between CLA-resistance and MALToma (P = 0.021).  et al. (2021) in Nanjing (1.83%). Nonetheless, the AMX-resistance rate was still the lowest among the four antimicrobials in this study. Such phenomena revealed less utilization of AMX in this area. AMX-based regimens might have the potential to be adopted as an empirical clinical treatment prescription in the future. MET-resistance rates differ all over the world. In western countries, it gained improvements after regulated uses, such as 16%, 25% and 27% respectively in Austria (Zollner-Schwetz et al., 2016), Portugal (Lopo et al., 2018) and Spain (Macias-Garcia et al., 2017). Meanwhile, in China, the use of inexpensive, easily obtained prescription medicine is easier to develop resistance (70%) (Chen et al., 2022). Within China itself, compared to other areas (Yu et al., 2019;Zhang et al., 2015b), the present study reported a high overall MET-resistance rate (79.14%). Though the typical empirical regimen(proton pump inhibitor, MET, CLA, and AMX) with increased dosage of MET or AMX might be an alternative to partly overcome the resistance to MET (Chen et al., 2019;Liang et al., 2013;Zhang et al., 2015a) , increased incidence of adverse side effects such as vomiting, nausea and rashes require consideration (Graham & Lee, 2015). Therefore, when prior susceptibility testing is not available, we suggest empirical treatment regimens containing MET be preferably abandoned. Resistance rate to LFX was up high to 41.10% in this study, higher than in China (31%), but almost similar to those in Shanghai (40.7%) (Yu et al., 2019) and Shenzhen (39.8%) (Lyu et al., 2020). Recently, LFX has been widely applied in H. pylori eradication treatment for the high CLA-and MET-resistance. A research discovered LFX-resistance were significant different between women and men (40.5% versus 21.5%), probably relating to the extent of use for urinary tract infections (Boyanova et al., 2012). We also found that the resistance to LFX presented a higher rate in female patients (27.61%) than that in male patients (13.50%). Such results suggested that we should reduce the frequency of LFX application among women population. Resistance to AOZ or TC was not detected in this study. Similar phenomenon has been observed in a population living in Sichuan Province, where the primary resistance rate of AOZ or TC was 0.8% (Tang et al., 2020). TC is not a common prescription in China, which means that semisynthetic TC derivatives such as minocycline might replace TC to gain good efficiency. Further randomized controlled trials need to be conducted. In terms of resistance patterns, among strains with no antibiotic resistance, treatmentnaïve strains accounted for a larger proportion. In other resistance patterns, treatmentexperienced group was the majority. Dual resistance was an exception, treatmentnaïve population (41.18%) were more frequently found than the other(29.45%). Dual resistance occupied the largest ratio (30.67%) of resistance patterns. Isolates exhibiting simultaneous resistance to CLA and MET had the largest prevalence (19.63%), followed by CLA+LFX(6.13%). Dual antibiotic resistance rate to CLA and MET from an Iran study (n = 12, 12%) were similar to ours and demonstrated the combination of those two antibiotics were not recommended as the first-line treatment (Hamidi et al., 2020). Based on our data, the frequency of triple antibiotic resistance was 25.77%, with LFX+MET+CLA predominating (n = 31, 19.02%). This might result in the failure eradication in this region and indicate advance drug susceptibility tests are required for the use of these drugs. Some factors were investigated to influence antibiotic resistance profiles in the present study. A significant relationship was found between CLA-resistance and endoscopic diagnosis (P = 0.010), we further analysed that MALToma (P = 0.021) was associated with resistance to CLA (Table 4). It has been reported that 60-80% of MALToma patients with H. pylori would be in remission after eliminating the bacteria (Nakamura & Matsumoto, 2013), and long-term CLA use or the immunomodulatory drug lenalidomide also work well (Raderer & Kiesewetter, 2020). An Australia study detected MALToma patients with H. pylori from 1997 to 2014 and concluded CLA-resistance rate was as low as 15%, which supported eradication regimens containing CLA (Bilgilier et al., 2016). However, the present high CLA-resistance rate and its close association with MALToma suggest us antimicrobial susceptibility tests should be required among those patients to select the appropriate antimicrobial eradication therapies or refined immunomodulatory strategies. We found a significant difference between MET resistance and age (P < 0.001). MET is a high prescription rate medicine (Ghotaslou et al., 2018). Choe et al. (2022) concluded that previous MET exposure caused a significant reduction in eradication rate of bismuth quadruple therapy. Another study (Lee et al., 2020) calculated that the odds ratio of past MET use leading to eradication failure was 3.468. Therefore, MET resistance might be related with an increased exposure of drugs. Antibiotics should be chosen more cautiously among patients of all ages. Other factors including gender, BMI, hypertension, diabetes mellitus and so on had no significant relationship with antibiotic resistance. The results of the above data might be due to the tiny sample size, which required to be further explored by expansion of the sample size.
Our study also had limitations. There were limited primary treatment subjects in this study. This is because in clinical practice, antibiotic sensitivity tests are usually applied to patients who have failed eradication several times. Patients without previous treatment history tend to choose empirical treatment regimen, so it is difficult to collect samples from first-treatment subjects, which is also a real problem. But we still offered an updated profile of H. pylori infectious patients in Liaoning. Researchers could carry this study continuously, by combining multiple hospitals and expanding the sample size, especially the number of subjects without experienced treatment history, and to explore a more stable antibiotic resistance profile in Liaoning. It is vital for clinical physicians to modify prescriptions, prevent unnecessary antibiotic consumption and evaluate the effectiveness of empirical treatment in time to improve eradication rate.

CONCLUSIONS
In conclusion, among this north China population, the primary resistance rates of LFX, MET, CLA, and AMX were relatively high and different in comparison with the overall or regional resistant data in China. To achieve good success rates, prior antimicrobial resistant tests are suggested before antibiotic prescription in H. pylori eradication to achieve good success rates and avoid severe antibiotic resistance.