The study showed the benefit of the bismuth add-on to 2-week clarithromycin-based triple therapy with propensity matching score analysis and investigated whether the bismuth add-on can overcome the clarithromycin resistance with arithmetic calculations indirectly. The study also showed two probiotics supplements (Lactobacillus casei variety rhamnosus and Bacillus Subtilis·Enterococcus faecium culture) might reduce the adverse event and enhance the eradication success.

Though the socioeconomic improvement in the twentieth century markedly decreased the prevalence of H. pylori infection in populations in developed countries, the need for treatment for H. pylori infection is still valid. To eradicate H. pylori in human, a variety of antibiotics and acid suppressants with enhanced potency have been tried at higher doses and for prolonged durations. However, we still have no definitive regimen of acceptable and liable eradication results. On top of that, along with the increased use of antibiotics globally, we encountered antibiotics-resistant H. pylori, especially to clarithromycin [15]. We are in urgent need to find a new drug that can increase the eradication effect without incurring resistance.

Bismuth is not new, but well known to be safe and used for long. Though it is not a kind of antibiotics, it has a bactericidal or bacteriostatic effect on H. pylori. Bismuth is commonly used as a component drug with other antibiotic combinations for H. pylori eradication [6, 10, 16, 17]. The bismuth-containing quadruple regimen were reported to result in more than 90% eradication success rate and to be effective even in cases of clarithromycin or metronidazole resistance. This effect was maintained even when tetracycline was substituted for amoxicillin [11].

In our study, authors sought to find the benefit of bismuth add-on by comparing clarithromycin-based triple regimens with and without bismuth. Bismuth add-on increased the eradication rate from 81.4% to 91.5% in ITT analysis (p-value = 0.023) and from 88.7% to 97.3% in PP analysis (p-value = 0.008). With bismuth add-on, the odd ratio for treatment failure decreased to 1 of 2.475 in ITT analysis and 1 of 4.875 in PP analysis. Our results look compatible and supportive with previous reports about the bismuth add-on. A recent European registry study reported an eradication success rate of 88% in intention-to-treat and 94% in per-protocol analysis when bismuth was added to a 14-day standard triple-drug regimen [18]. A retrospective single-arm study by Kim et al. reported that when it was added to the 2-week standard triple regimen the final eradication success rate reached 77.1% in the subjects with clarithromycin resistance [19]. This was also a better outcome than that of conventional clarithromycin-based triple regimen for clarithromycin-resistant H.pylori. Horie et al. reported eradication rates of 88% and 45% with clarithromycin-based triple regime for clarithromycin-sensitive and clarithromycin-resistant cases, respectively [20].

To the question of whether the bismuth add-on can overcome the clarithromycin resistance, the answer was not directly explored in this study. However, the possibility can be estimated by arithmetic calculations using relevant data [20, 21] and the authors’ study outcomes. If we cite the recent Korean research result [21] on the clarithromycin resistance, 17.8%, and eradication outcomes from Horie’s study [20], the calculated estimation of eradication success with conventional triple therapy (PAC) in Korean population is 80.3%. This simple calculation result, 80.3%, is compatible with our result of 81.4% (95% CI 0.7423–0.8849) of eradication success rate in ITT analysis for PAC group and supportive of the reliability of our study. The effect of the bismuth add-on on the clarithromycin-resistant H. pylori can also be estimated by following arithmetic calculation; [91.5% of final eradication rate in ITT analysis of PACB group = prevalence of clarithromycin-resistant H. pylori infection (17.8%) x unknown eradication rate (x) + prevalence of clarithromycin-sensitive H. pylori infection in Korea (82.2%) x unknown eradication rate (y)]. With this arithmetic calculation, the “y” value is calculated in the range of 88%(20)–100%, and “y” should not be less than “x”, then consequently the range of the “x” value should be between 52.4% and 91.5%. If the proportion of clarithromycin-resistant H. pylori increases, the lowest estimate of the interval will also increase. This is significantly higher than the 45% result of Horie et al. and is consistent with the results of other previous studies [18, 19, 22]. We can conclude that the bismuth add-on can overcome the clarithromycin resistance. Nevertheless, the authors do not advocate the use of a bismuth-added clarithromycin-based triple regimen for clarithromycin-resistant H.pylori based on the results of this study. Because it is reasonable to use metronidazole if clarithromycin resistance is known. The authors would like to emphasize that the bismuth add-on can improve the overall eradication outcomes of a 2-week clarithromycin-based triple regimen in situations where an empirical treatment is required. Contrary to the results of this study, Wu et al. reported that the eradication rate could not be improved even with a bismuth add-on for the clarithromycin-based triple regimen [23]. However, Wu et al.'s study differ from the authors' study in that it was the result of administering a clarithromycin-based triple regimen for one week in an empirical condition.

One of the concerns, when we use bismuth, is compliance and possible adverse events. In the authors’ study, the proportion of subjects who responded to take over 85% of the total doses was over 98% in both PACB and PAC groups. It looks higher than that of other reports even including prospective studies in which compliance was reported around 81–98% with the standard triple regimen [24, 25]. We admit the concern about recall bias due to the retrospective design of this study. However, the authors would like to suggest another possible explanation for better compliance than expected. All subjects in this study were prescribed two probiotics supplements along with antibiotics. Probiotics including Lactobacillus rhamnosus GG are known to help to reduce the side effects of eradication therapies and improve compliance with therapy [26]. A meta-analysis by Dang et al. reported that higher eradication rate in probiotics supplementation groups than in control (ITT analysis: RR 1.122, 95% CI 1.086–1.159, PP analysis: RR 1.114, 95% CI 1.070–1.159) and lower incidence of adverse event (RR 0.735, 95% CI 0.598–0.902) [27]. Argues are still on the table about the beneficial strains of probiotics and the definite mechanism for these effects, a variety of probiotic strains have been listed to be helpful in H. pylori eradication [28].

Regarding the adverse events, over 35% of subjects reported having discomfort or adverse events during administration. The authors acknowledge the concerns and potential for recall bias and missing records when collecting adverse event data. However, again, the authors infer that the coadministration of probiotics supplements might play a role in reducing the occurrence of possible adverse events. More importantly, the limitations of the retrospective investigation were the same for both PAC and PACB groups, and there was no difference in the incidence of adverse events according to the bismuth add-on or not. Almost all subjects in PACB group experienced dark colored stool, but the change of stool color predicted though the explanation before taking the regimen was not judged to be uncomfortable. Our results were comparable to other previous prospective studies using bismuth [25, 29] and a study from China reported 28% of adverse event incidence with 2-week PACB regimen [1]. Actually tetracycline was suggested as a major reason for adverse events in the bismuth-containing quadruple regimen[24] and bismuth itself was reported to reduce the discomfort during eradication therapy [30, 31].

Regarding rescue therapy after 1st line standard triple regimen failure, recent guidelines recommend bismuth-containing quadruple regimen as the 2nd line therapy. The authors' institution uses bismuth-containing quadruple regimen as recommended or metronidazole-based triple regimen as a 2nd line if the 1st line was clarithromycin-based triple regimen. This is most likely due to clarithromycin resistance when clarithromycin-based triple regimen therapy fails, and because metronidazole-based triple regime is generally judged to be relatively convenient to take for bismuth-containing quadruple regimen. A total of 16 of the enrolled subjects were diagnosed with treatment failure after UBT examination. The results of secondary treatment for them will be further analyzed in the future.

This study has some limitations. First, due to the retrospective study design, the concern about recall bias and missed records was inevitable. However, an important issue in attempting randomized studies to measure the effects of bismuth add-on is the ethical question of the probably expected negative effects in a control group. Recently in Korea, there have been significant changes in the usage direction of bismuth. From January 2019, under ‘Ministry of Health and Welfare Drug Notification No. 2018–174 on the standards of medical care benefits and Detailed information on the application methods’, bismuth drug can be used without restrictions to eradicate H. pylori beyond the previously approved indications. Thanks to this situational condition, the authors could carry out a parallel comparative study with and without bismuth. In this study, the two comparison groups had nearly identical baseline characteristics regarding residence area and time of intervention. The authors adopted propensity score matching analysis to minimize possible bias in outcomes by undetected factors. Second, this is a single-centered investigation. We need a multi-centered investigation to generalize our results. Third, antibiotics susceptibility test was not employed in this study. The study design was determined by considering that an empirical therapy is performed in most clinical practices, but the authors acknowledge that the interpretation of the study results would be more precise if the information on antibiotic susceptibility were available.References: As per pubmed findings, citation details [DOI] for Reference [8, 24] have been inserted. Kindly check and confirm the inserted details.Yes, I agree with the amendment.