The use of real-world evidence to generate cost analysis of antibiotic susceptibility testing (AST) in patients with treatment failure in Thailand: A large population-based study.

1
Introduction
Helicobacter pylori (H. pylori), a Gram-negative, rod-shaped bacterium, regularly causes chronic gastric mucosal inflammation. H. pylori infection results in a wide range of diseases from asymptomatic chronic gastritis, peptic ulcer disease, to gastric cancer [1]. Over 4 billion people worldwide had H. pylori infection [2]. The prevalence of H. pylori infection varied among countries from 18.9 % to 87.7 % [2]. Approximately one third of Thai population were infected with H. pylori as reported in the nationwide study [3]. Early diagnosis and eradication of H. pylori infection are recommended to reduce risk of gastric cancer [4,5]. Treatment goal for H. pylori infection is achieving >90 % eradication rate [6]. However, the cure rate of standard triple therapy as first-line regimen in Thailand has been decreasing to 80 % [7]. One of the most important barriers to treatment success is rising antibiotic resistance [8]. As clarithromycin is frequently included in the first-line regimen composition, its resistance has a significant impact on eradication success. Therefore, susceptibility test is important to determine regional antibiotic resistance patterns to develop the most effective therapy in an area with high levels of antibiotic resistance [4].
Antimicrobial susceptibility testing (AST) is an in vitro test to determine activity of antibiotics that can effectively inhibit the growth of bacteria. The previous recommendation stated that H. pylori culture and AST should be performed after treatment failure [9], while the current guideline recommends that AST should be routinely done even before first-line treatment to prevent antibiotic overuse and promote antibiotic stewardship [4]. However, several issues have to be considered when performing culture and AST including test availability, proper specimen handling, microbiological expertise, and costs. Prior studies evaluating cost-effectiveness between susceptibility-guided treatment and empirical therapy demonstrated inconsistent results. Although these studies reported significantly higher eradication rates in tailored treatment group than empirical treatment group for first-line eradication [[10], [11], [12]], tailored treatment was cost-effective in two studies [11,12]. For refractory H. pylori infection, one previous study reported higher cost when using genotypic resistance-guided therapy compared to empirical therapy without difference in eradication rates [13].
Nowadays, a rising number of patients experience H. pylori treatment failure because of increasing antibiotic resistance. AST is recommended after treatment failure [7]. Until now, there have been insufficient data regarding eradication rates of therapies after treatment failure in Thailand. This study aimed to determine eradication rates of susceptibility-guided therapy and empirical treatment in patients with refractory H. pylori infection in Thailand. Moreover, this study also performed cost analysis between AST-guided therapy and empirical treatment.

2
Methods
2.1
Study design
This was a retrospective cohort study conducted at tertiary care center in Thailand. Retrospective chart review between March 6, 2014, and October 31, 2021 was performed. This study included patients aged >15 years with refractory H. pylori infection defined as positive rapid urease test, histopathology, or culture after completing one or more eradication regimen(s). Patients’ demographic data, comorbidities, endoscopic results, H. pylori culture and AST, eradication regimens, and treatment cost were extracted from the medical database and extensively reviewed. H. pylori treatment at each follow-up visit was recorded. The exclusion criteria were patients who did not have confirmatory testing for H. pylori eradication.
The primary aim of this study was to determine eradication rates of susceptibility-guided therapy and empirical treatment in patients with treatment failure. Secondary outcomes were to investigate current antimicrobial resistance patterns and perform cost analysis comparing susceptibility-guided therapy with empirical treatment.

2.2
H. pylori culture and antimicrobial susceptibility testing
If a rapid urease test yielded a positive result, H. pylori culture would be subsequently performed. Two antral biopsies in an Eppendorf tube were ground in the broth. A sterile inoculating loop was used to streak culture on a Mueller Hinton II agar medium. After incubation at 37 °C in a microaerophilic atmosphere (5 % O2, 10 % CO2, and 10 % H2) for 3–5 days, colony formation of H. pylori is presented on the agar plate [14]. Fresh H. pylori colonies (<72 h of growth) at an inoculum concentration of 3 McFarland standards were used to perform the Epsilometer test (E-test). At 3–5 days after inoculation, the E-test strip with each antibiotic was examined to determine the minimum inhibitory concentrations (MICs) defined by the point of intersection between the inhibition ellipse and scale on the E-test strip [15]. According to the European Committee on Antimicrobial Susceptibility Testing (EUCAST), resistant strains were defined by MIC values of >0.125 mg/L for amoxicillin (AMX), >0.5 mg/L for clarithromycin (CLR), >8 mg/L for metronidazole (MTZ), >1 mg/L for levofloxacin (LVX), and >1 mg/L for tetracycline (TET) [16].

2.3
Definitions
Refractory H. pylori infection or H. pylori treatment failure was defined as persistent H. pylori infection after one or more treatment regimen(s) completion. Persistent H. pylori infection was diagnosed by a positive result of either one of tests as follows: rapid urease test, histopathology, or culture [8].
Multidrug-resistant strains (MDR) was defined as strains which develop resistance to at least one antibiotic in three or more classes of antibiotics [17].
Triple therapy consisted of proton pump inhibitor (PPI), amoxicillin (1000 mg bid), and clarithromycin (500 mg bid) for 14 days.
Concomitant therapy consisted of PPI, amoxicillin (1000 mg bid), metronidazole (500 mg bid), and clarithromycin (500 mg bid) for 10 days.
Sequential therapy consisted of PPI and amoxicillin (1000 mg bid) for 5 days followed by PPI, metronidazole (500 mg bid), and clarithromycin (500 mg bid) for 5 days.
Quadruple therapy consisted of PPI, bismuth subsalicylate (1048 mg bid), metronidazole (400 mg tid), and tetracycline (500 mg qid) for 14 days.
Tailored therapy or susceptibility-guided therapy was defined as an eradication regimen comprising antisecretory drug and susceptibility-guided antibiotics. Patients without AST performed were considered using empirical therapy.
Early AST was defined as AST performed immediately after first-line treatment failure.
Nonadherence was defined as taking less than 80 % of prescribed medication for H. pylori treatment (e.g., taking medication for <11 days of 14-day regimen) [18].

2.4
Statistical analysis
All data were analysed by using SPSS version 22 (SPSS Inc., Chicago, IL, USA). Chi-square test, or Fisher's exact test were performed for analysis of categorical variables, whereas Student's t-test was used for continuous variables and reported as mean ± standard deviation (SD). After using univariate analysis to identify factors related to achieving successful eradication after treatment failure, every variable with p-value <0.05 was further analysed by multivariate analysis to determine an adjusted p-value. All tests were two-sided and statistical significance was defined as p-value of <0.05.

2.5
Ethics statements
The Human Research Ethics Committee of Thammasat University (Medicine), Thailand (MTU-EC-IM-4-217/66) provided ethical approval for this study. The authors conducted this research according to the good clinical practice guideline, as well as the Declaration of Helsinki. All data had been fully anonymized before they were accessed. The Ethics Committee waived the requirement for informed consent because of no greater than minimal risk for study subjects. The authors had no accessibility to data that could identify individual participants during or after data collection.

3
Results
3.1
Demographic data of patients with H. pylori treatment failure
There were 1080 patients diagnosed with H. pylori infection during esophagogastroduodenoscopy at tertiary care center in Thailand. Of 1080 patients, 315 (29.2 %) had H. pylori eradication failure [140 men and 175 women; mean age 58.4 ± 13.1 (range 23–89) years]. Similar baseline characteristics between groups of treatment success and treatment failure were demonstrated in Table 1. The most common first-line regimen was triple therapy (40.2 %), followed by concomitant therapy (14.3 %) and bismuth quadruple therapy (14.3 %). Eradication rates of first-line regimens were as follows: triple therapy (68.1 %), concomitant therapy (74.7 %), sequential therapy (59.5 %), bismuth quadruple therapy (68.2 %), levofloxacin triple therapy (75.4 %), and vonoprazan-containing therapy (78.3 %). In multivariate analysis, sequential therapy was significantly associated with treatment failure [OR 1.66 (95%CI 1.01–2.74), p = 0.045)], whereas vonoprazan-containing therapy was associated with treatment success [OR 1.60 (95%CI 1.04–2.48), p = 0.034)]. Medication nonadherence [OR 37.97 (95%CI 8.97–160.65), p < 0.001)] contributed to treatment failure. Medication nonadherence resulted from intolerance to adverse effects such as nausea and vomiting (19/30, 63.3 %), forgetfulness (9/30, 30 %), and drug allergy (2/30, 6.7 %). Patients with treatment failure were subsequently classified into two groups. One group had no AST done and received empirical therapy (n = 230), while the other had AST done and used susceptibility-guided therapy (n = 85) as reported in Table 2. Baseline characteristics and first-line regimens between two groups were comparable. One patient in empirical therapy group and five patients in AST-guided therapy group were lost to follow-up and excluded before calculating eradication rates. The flow diagram of patients in this study was demonstrated in Fig. 1. The group of AST-guided therapy had significantly higher eradication rate than the group of empirical therapy (97.5 % vs. 65.5 %, OR 20.54; 95%CI 4.92–85.81, p < 0.001). For subgroups of AST-guided therapy, eradication rates between early AST and AST after treatment failures were not different (95.7 % vs. 98.2 %, p = 0.49). The most commonly used empirical therapy for second-line treatment was levofloxacin triple therapy (44.1 %), followed by bismuth quadruple therapy (28.8 %), and clarithromycin-containing regimens (24 %). Eradication rates of empirical treatment were as follows: levofloxacin triple therapy (69/101, 68.3 %), bismuth quadruple therapy (44/66, 66.7 %), clarithromycin-containing regimens (triple therapy 17/29, 58.6 %; concomitant therapy 12/16, 75 %; sequential therapy 3/10, 30 %). There were 8.3 % of patients in the empirical therapy group with medication nonadherence during first-line treatment. The multivariate analysis demonstrated that medication adherence was an only one factor associated with successful eradication (OR 9.84; 95%CI 1.12–86.11, p = 0.039) in empirical therapy group. All second-line regimens except for quinolone-based therapy were not different between empirical therapy group and early AST group as demonstrated in Supplementary Table 1. The proportion of quinolone-based therapy in early AST group was higher than empirical therapy group (69.6 % vs. 44.1 %, p = 0.02).

3.2
Antimicrobial susceptibility testing in patients with H. pylori treatment failure
Patients with H. pylori treatment failure were classified into three groups: 1) no AST (n = 230), 2) early AST (n = 24), and 3) AST after treatment failures (n = 61). AST was performed in 85 patients as demonstrated in Table 3. Most of strains (84.7 %) were resistant to antibiotics, while only 15.3 % were susceptible strains. Resistance rates to levofloxacin (LVX), metronidazole (MTZ), clarithromycin (CLR), and amoxicillin were 57.6 %, 51.8 %, 44.7 %, and 4.7 %, respectively. No tetracycline resistance was observed in this study. The group of AST after multiple treatment failures had significantly higher resistance of LVX (68.9 % vs. 29.2 %, p = 0.001), CLR (52.5 % vs. 25 %, p = 0.022), and MDR strains (32.8 % vs. 8.3 %, p = 0.020) than the early AST group. However, MTZ resistance was high in both early AST and AST after treatment failures groups, while amoxicillin resistance were low in both groups. The most common MDR organisms were CLR, LVX, and MTZ-resistant strains (21.2 %) which were only detected in the group of AST after treatment failures.

3.3
Cost of H. pylori eradication in patients with treatment failure
Total cost of H. pylori eradication in patients with treatment failure was calculated as the sum of medication costs, posttreatment testing costs, hospital visit costs and AST if performed. Medication cost was calculated as the sum of cost of antisecretory drugs (PPI or potassium-competitive acid blocker) and antibiotics multiplied by duration of treatment. Urea breath test (UBT) was regularly used as posttreatment testing for refractory H. pylori infection. UBT cost was $81.1 [2700 Thai Baht (THB)]. The cost of H. pylori culture and AST was $75.1 (2500 THB). The average total cost and all subcategory costs of H. pylori eradication per patient were illustrated in Table 4 and Fig. 2. The mean cost of first-line medication of the group without AST, early AST, and AST after treatment failures were not different from each other ($46.8 vs. $42.6 vs. $44.1, respectively, p > 0.05). Costs of first UBT and first hospital visit were also similar among three groups resulting in comparable total cost of first-line treatment. The cost of H. pylori culture and AST was added in the group of early AST and AST after treatment failures. The mean cost of subsequent medication in the early AST group was significantly lower than the group without AST ($28.2 vs. $84.3, p < 0.001) and AST after treatment failures ($28.2 vs. $198.0, p < 0.001). The mean medication cost increased after every failed eradication as depicted in Fig. 3. The mean cost of subsequent UBT in the early AST group was also lower than the group without AST ($81.1 vs. $122.2, p < 0.001) and AST after treatment failures ($81.1 vs. $261.1, p < 0.001). The mean cost of subsequent hospital visits in the early AST group was also lower than the group without AST ($45.0 vs. $52.6, p < 0.001) and AST after treatment failures ($45.0 vs. $78.4, p < 0.001). Consequently, early AST group had significantly lower total average cost of H. pylori eradication than the group without AST ($368.2 vs. $402.0, p = 0.034) and AST after treatment failures ($368.2 vs. $752.8, p < 0.001). The early AST group had the lowest cost of subsequent medication, posttreatment UBT, and hospital visits.

4
Discussion
H. pylori eradication confers protection against gastric cancer. Nowadays, current first-line regimens provide eradication rates of less than 80 % owing to rising clarithromycin resistance [19]. Treatment failure becomes a challenging problem that needs to be solved. This study demonstrated that patients with treatment failure using susceptibility-guided therapy had significantly higher eradication rate than the group using empirical therapy. Furthermore, cost analysis also reported that the group with early AST had lower total average cost of H. pylori eradication than the group without AST and AST after multiple treatment failures. Early AST might be a cost-effective strategy which could enhance the eradication rate of refractory H. pylori infection.
Treatment failure has become a major concern in management of H. pylori infection in recent years. Eradication failure rate after first-line therapy in our study (29.2 %) was comparable to others (20.2%–23.4 %) [[20], [21], [22]]. Majority of patients in this study (40.2 %) and others (39.4%–80.2 %) from different global regions used clarithromycin triple therapy as first-line regimen despite high prevalence of clarithromycin resistance in all regions including Asia, Europe, and North America [[20], [21], [22]]. This could be due to discrepancy between real-world practice and guideline recommendation. Although gastroenterologists in our study followed the latest Thai guideline in 2015 which recommended triple therapy as first-line treatment [7], this study highlights that the national guideline will require revision in the near future according to declining cure rates of first-line clarithromycin-containing regimens. The reason why sequential therapy was significantly associated with treatment failure might be due to regimen complexity and high dual clarithromycin and metronidazole resistance. In contrast, vonoprazan-containing therapy as first-line regimen was a factor related to treatment success. This might be because of vonoprazan's properties as a potent acid blocker and its superior efficacy to proton pump inhibitor-based therapy for eradication of clarithromycin-resistant strains [23]. Treatment failure occurs as a consequence of either host or bacterial factors [8]. Medication nonadherence was associated with eradication failure in this study which was similar to the prior study [24]. However, it might be difficult to improve the adherence rate in this study since the adherence rate in treatment failure group was already high (91.1 %). The most common cause of medication nonadherence in this study was intolerance to side effects from antibiotics. Probiotics such as Lactobacillus spp and multiple strains might be beneficial for reducing antibiotic-related side effects which could further improve eradication rates [25]. Apart from patient nonadherence, antibiotic resistance is also a principal cause of treatment failure [8]. Although there were approximately 20 % of patients with AST, this study revealed higher proportion of MDR strains in treatment failure group than successful treatment group. Our important data about local antibiotic resistance patterns and previously failed regimens could assist in appropriate regimen selection after treatment failure.
Antimicrobial susceptibility testing is a valuable tool to guide antibiotic choice for refractory H. pylori infection. After treatment failure, this study revealed that AST-guided therapy provided significantly higher eradication rate than empirical therapy (97.5 % vs. 65.5 %). The recent meta-analysis reported no significant difference in eradication rates for rescue therapies between susceptibility-guided therapy and empirical treatment; however, there was considerable heterogeneity of studies between groups (RR 1.07; 95%CI 0.97–1.18; I2 78 %) [26]. Most of empirical therapies prescribed for second-line treatment in this study were levofloxacin triple therapy and bismuth quadruple therapy yielding low cure rates (65–70 %). High level of LVX resistance (57.6 %) could explain markedly reduced second-line eradication rates. This study also demonstrated that CLR-containing regimens yielded low eradication rates (30–75 %) and should not be used as empirical therapy after treatment failure. Moreover, this study reported that resistance to LVX, CLR and MDR substantially increased after multiple treatment failures compared with failure after only one regimen. This data emphasized that repeated antibiotic exposure after each time of eradication could develop more resistant H. pylori strains. A similar result was observed in the previous study which demonstrated extremely high secondary resistance of CLR, MTZ, and LVX up to 50–80 %, whereas amoxicillin and tetracycline resistance remained low [27]. From AST results in our study, it can be implied that early use of AST-guided therapy not only resulted in early successful eradication, but also prevented the development of more resistant strains in the latter eradication.
Availability and the cost of AST are essential issues to take into consideration to generalize the use of AST in H. pylori treatment failure. This study demonstrated that early AST group had lower total average cost of eradication than the group without AST and AST after treatment failures. This was because early AST group received tailored therapy since second-line treatment and most of them achieved successful second-line eradication. Therefore, they had the lowest cost of medication after first-line treatment failure, posttreatment UBT, and hospital visits. Most of previous studies were conducted to compare cure rates of first-line treatment between tailored regimen and empirical treatment which was different from our study evaluating eradication rates after treatment failure [[10], [11], [12]]. All studies exhibited higher eradication rates of tailored treatment, while the cost was lower or comparable to empirical therapy [[10], [11], [12]]. One study also mentioned that tailored therapy not only improved the eradication rate (91.7 % vs. 82.6 %) but also achieved a cure with shorter treatment duration (79.8 vs. 99.2 days), reduced adverse events (12.9 % vs. 14.8 %), and lower total medical costs than empirical group for first-line treatment [28]. Similarly, our study demonstrated higher eradication rate and lower cost of treatment, but we highlighted this in subsequent treatment beyond first-line therapy. One previous study displayed different results from our study reporting higher cost of genotypic resistance-guided therapy ($6920 to additionally cure one patient) compared to empirical therapy but no difference in eradication rates (78 % vs. 72.2 %) [13]. This might be because of different study design and population. The study by Liou et al. was a randomized trial evaluating patients with at least 2 failed eradications, while our study assessed patients who failed first-line treatment. Furthermore, Liou et al. study demonstrated considerably higher antibiotic resistance rates (resistance to CLR 90.1–95.4 %, LVX 58.6–62.1 %, MTZ 63.5–68.2 %, AMX 13.8 %) than our study [13]. This might be implied that patients having more failed treatments would develop more resistant strains causing more difficult-to-treat refractory H. pylori infection [29].
Our study has several strengths. This was a large population-based study reporting better eradication rates of susceptibility-guided therapy than empirical therapy in patients with H. pylori treatment failure. Moreover, this research highlighted that performing early AST might be more cost-effective than AST after multiple treatment failures. However, this study also had some limitations. Firstly, there were some missing data of AST due to the nature of retrospective cohort study. We tried our best to collect as many positive culture results as possible in our analysis. Secondly, healthcare and medication costs substantially vary among countries. Moreover, AST for H. pylori is not available in some areas in Thailand. Therefore, this might limit generalizability in areas with different healthcare costs or lack of AST.
In conclusion, early AST is cost-effective for treatment of H. pylori eradication failure. Susceptibility-guided therapy provided significantly higher eradication rate than empirical therapy in patients with treatment failure. The most appropriate first-line treatment depends on local antibiotic resistance pattern. Vonoprazan-containing regimen was associated with treatment success and might be used as first-line regimen. Medication adherence should be enhanced since first-line treatment to prevent eradication failure.

CRediT authorship contribution statement
Natsuda Aumpan: Writing – review & editing, Writing – original draft, Visualization, Validation, Software, Resources, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Pornpen Gamnarai: Writing – review & editing, Resources, Investigation. Arti Wongcha-Um: Writing – review & editing, Validation, Methodology, Investigation. Muhammad Miftahussurur: Writing – review & editing, Supervision, Methodology. Yoshio Yamaoka: Writing – review & editing, Validation, Supervision, Methodology, Funding acquisition. Ratha-korn Vilaichone: Writing – review & editing, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Data availability statement
The supplementary data have been deposited at Figshare (https://figshare.com) with accession numbers (https://doi.org/10.6084/m9.figshare.26381728.v1).

Ethics Declarations
This study was reviewed and approved by the Human Research Ethics Committee of Thammasat University (Medicine), Thailand, with the approval number: MTU-EC-IM-4-217/66. Informed consent was not required for this study because of no more than minimal risk to study subjects.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.