A randomized, multicenter and noninferiority study of amoxicillin plus berberine vs tetracycline plus furazolidone in quadruple therapy for Helicobacter pylori rescue treatment

Objective Helicobacter pylori (H. pylori) infection is closely associated with gastric ulcers and gastric adenocarcinomas. We aimed to assess the efficacy and safety of a quadruple regimen with amoxicillin plus berberine vs tetracycline plus furazolidone in rescue therapy for H. pylori eradication. Methods We conducted a randomized, open‐label, multicenter, noninferiority trial. Patients with previous treatment failures recruited from five centers were randomized (1:1) to receive a regimen with esomeprazole and bismuth plus either berberine and amoxicillin (the BA group) or tetracycline and furazolidone (the TF group) for 14 days. Their H. pylori infection status was confirmed 4‐8 weeks after treatment. The primary outcome was the eradication rate. The secondary outcomes included the rates of symptom improvement, compliance, and adverse events. This study was registered at ClinicalTrials.gov (NCT03609892). Results Altogether 658 participants were consecutively enrolled. An intention‐to‐treat analysis demonstrated that the two regimens achieved a similar eradication rate (76.3% vs 77.5%; P = 0.781). The per‐protocol analysis reached a similar result (81.5% vs 85.0%; P = 0.278). The eradication rate reached in the BA group was greater than the pre‐established margin of noninferiority, at −10% (the lower bounds of the 95% CI were −7.66% and −9.43%, respectively). The rate of adverse events was lower for the BA group than the TF group (18.5% vs 26.1%, P = 0.024). Rates of compliance and symptom improvement were similar for the two therapies. Conclusion The efficacy of both regimens in rescue treatment for H. pylori eradication was satisfactory, 14‐day BA‐based quadruple therapy is noninferior to the TF‐based therapy.


| INTRODUCTION
It is generally accepted that Helicobacter pylori (H. pylori) infection plays a vital role in the development of chronic gastritis, gastric and duodenal ulcers, and gastric malignancy. 1 Eradicating H. pylori in infected individuals help in the resolution of gastritis and upper gastrointestinal ulcers and may reduce the occurrence of gastric cancer. [2][3][4] Thus, standard eradication therapy should be recommended to all patients with a known H. pylori infection. 5 However, a continuous increase in bacterial antibiotic resistance is a huge challenge for H. pylori eradication. Unfortunately, the eradication efficacy of the classical standard triple therapy has been disappointing (being less than 80% for decades). 6,7 Resistance to the main antibiotics, such as clarithromycin and levofloxacin, in H. pylori eradication treatment has soared in recent decades. 8 Previous studies have reported that clarithromycin resistance rates have soared to more than 30% in Japan and Italy, and more than 40% in Turkey, although those in Sweden and the Taiwan region remain relatively low (about 15%). 9 The Fifth Chinese national consensus report on the management of H. pylori infection has recommended seven antibiotic combinations that include six kinds of antibiotics. 10 Nevertheless, according to recent data from China, primary H. pylori resistance rates to the main antibiotics (clarithromycin, levofloxacin, and metronidazole) were 20%-50%, 20%-50% and 40%-70%, respectively, which exceeded our expectations. [11][12][13][14][15][16][17] Quadruple therapies comprising two kinds of antibiotics, such as furazolidone, tetracycline, and amoxicillin, along with proton pump inhibitors and bismuth, have been proven to reach a relatively satisfactory rate (> 90%) for those with high resistance to the other main antibiotics. 18,19 However, due to the high unavailability of furazolidone and tetracycline, researchers are searching for treatment regimens that are more effective and safer.
There have been several studies on the efficacy of traditional Chinese medicine for treating H. pylori infection. Some of them have found that certain kinds of herbs exert an excellent inhibitory effect on H. pylori. 20,21 As a traditional Chinese medicine, berberine is derived from herbs and is widely used to treat different digestive diseases, such as diarrhea and bacterial gastroenteritis. 22 In addition, berberine is very inexpensive (0.2 Yuan [RMB]/tablet) in China and can easily be obtained over the counter. Recently, it has been reported that berberine can significantly lower the minimum inhibitory concentrations of tetracycline and amoxicillin against H. pylori strains, through the mechanism of decreasing the expression of hefA mRNA. 23 Berberine has synergistic effects with some common antibiotics, suggesting that it could be useful when combined with other antibiotics for the eradication of antibiotic-resistant H. pylori strains. 24 Moreover, the inhibitory effect of berberine on H. pylori strains on patients with peptic ulcers has been tested in some pharmacological studies, showing that berberine can inhibit the activity of arylamine N-acetyltransferase, thus inhibit the proliferation of H. pylori. [25][26][27] We previously performed two prospective randomized controlled trials on the effect of berberine for H. pylori eradication. Our results verified that a berberine-containing therapy (berberine, esomeprazole, amoxicillin and clarithromycin) was not inferior to the first-line standard quadruple regimen (bismuth, esomeprazole, clarithromycin, and amoxicillin) in the local region (ClinicalTrials.gov ID: NCT02296021). 28 We also reported that a bismuth-containing therapy including berberine plus amoxicillin was not inferior to that including amoxicillin plus clarithromycin in the first-line H. pylori eradication treatment (ClinicalTrials.gov ID: NCT02633930). 29 These results support the idea that berberine may be effective against H. pylori resistance. To explore the clinical application of berberine, we conducted this multicenter, randomized, controlled, noninferiority study on the efficacy and safety of a brand-new quadruple regimen including amoxicillin plus berberine in H. pylori rescue treatment for patients who had previously failed eradication therapy at least once.

| Participants
Participants infected with H. pylori in whom eradication treatment had failed at least once, such as standard triple and quadruple therapies with or without bismuth, were consecutively recruited for our study. Eligibility criteria were as follows: (a) adult participants aged between 18-70 years and of either sex; (b) patients in whom H. pylori eradication therapy had failed during the previous 2 years and had discontinued eradication therapy for more than 2 months; and (c) women who were not pregnant or lactating (child-bearing women were asked to use a medically acceptable contraception during the study period and the subsequent 30 days after the study.

| H. pylori infection detection
Detection of H. pylori infection was performed by using the 13

| Procedure
The participants were randomly allocated to receive a 14-day BAbased or TF-based quadruple regimen. Each participant's characteristics and medical history were obtained at enrollment. We also collected data on the occurrence of their clinical symptoms, including vomiting, nausea, abdominal pain, flatulence, diarrhea, constipation, early satiety, belching, hiccup, dysgeusia, headache, dizziness, drowsiness, fever, acid reflux, and heartburn. The severity of symptoms was categorized from 0 to 3 (0, 1, 2, and 3, represent none, mild, moderate, and severe, respectively). The frequency of symptoms was also categorized from 0 to 3 (0, 1, 2, and 3, represents none, mild, moderate, and severe, respectively). A total score of each symptom was calculated as follows: total score = frequency score × severity score.
The participants were required to come back to the hospitals after they had completed treatment for an assessment of their compliance, adverse events and symptom improvement. The participants were informed of the common adverse events and were instructed to record these events on a predesigned case report form. A four-point scale, which included none, mild (bothersome discomfort that did not interfere with normal life), moderate (sufficient discomfort to cause interference with normal life), and severe events (discomfort potentially meriting discontinuation of the therapy), was designed to evaluate adverse events prospectively. Participants were asked to stop using proton pump inhibitors and histamine-2 receptor antagonists for at least 14 days before any examination. The technicians carrying out the H. pylori tests were blinded to the therapies the participants received. The whole flow diagram of the study is shown in

| Outcomes
The primary outcome was the efficacy of H. pylori eradication treatment, as evaluated at least 4 weeks after the completion of the therapy, and the secondary outcomes included the overall rates of adverse events, improvement in clinical symptom (defined as improved when the total clinical symptom score was reduced by ≥ 50%, as evaluated at the 2nd week of the treatment and at 4 weeks after completion of the therapy) and participants' compliance. Patients who took less than 80% of all medications were determined to have a poor compliance after counting unused medications at completion.

| Statistical analysis and calculation of sample size
The efficacy of the quadruple regimen using tetracycline and furazolidone for H. pylori eradication has been reported to range from 75% to 90% for H. pylori rescue therapy. [30][31][32] We assumed we wouldachieve an 80% eradication rate in both groups, a noninferiority margin of δ = −10% with a β power of 90% and an α of 0.05 (one-sided). Based on the H. pylori eradication rate we assumed, the per treatment regimen required at least 274 participants to indicate noninferiority between the two groups. A 20% loss to follow-up rate was taken into consideration, and we expected to recruit at least 329 participants for each group (658 participants in total) for this trial. BA therapy was considered noninferior to TF therapy if the lower bound of the 95% confidence interval (CI) was higher than −10%; otherwise, noninferiority was not concluded. P value of less than 0.05 was considered statistically significant.  Table 1, there were no statistical differences in patients' baseline data, clinical characteristics or the antibiotics used in previous treatment regimens between the two groups. Moreover, nine participants had a positive RUT result or histological confirmation before treatment, and we obtained their endoscopic and pathological results on enrollment (two participants were diagnosed with ulcers, five with chronic atrophic gastritis, and the other two with chronic atrophic gastritis with intestinal metaplasia).  Table 2, according to the ITT analysis H. pylori was successfully eradicated in 76.3% of patients in the BA group and in 77.5% of those in the TF group. On PP analysis, the eradication rate reached 81.5% and 85.0% in the BA and TF groups, respectively. There was no statistical difference between the two regimens (P = 0.781 and 0.278 for the ITT and PP analysis, respectively). Moreover, the lower bound of the 95% CI for the eradication rate difference in the BA therapy was greater than the pre-established marginal value of noninferiority of −10% (−7.66% and − 9.43%, respectively). Metronidazole + levofloxacin 1 0

| Rates of adverse events, compliance, and symptom improvement
As shown in Table 3

| DISCUSSION
Due to a rapid increase in antibiotic resistance worldwide, the overall global eradication rate of H. pylori using the standard triple regimen has dropped to less than 80%. 33 While genetic polymorphisms in CYP2C19, unsatisfactory compliance, and the bacterium itself are considered the reasons for this phenomenon, antibiotic resistance is regarded as the most important factor, especially in China. 17 A study in Shanghai reported that H. pylori resistance rose from 8.6% to 20.7% for clarithromycin and from 10. on PP analysis. 35 However, due to adverse events and the lack of their availability, there is a great limitation in the use of tetracycline and furazolidone in the treatment of H. pylori infection.
Berberine is isolated from Coptis chinensis and is known for its antiprotozoal, antibacterial, and anti-inflammatory abilities. Studies have shown that berberine may effectively inhibit the growth of H. pylori, with a zone of inhibition reaching 25 mm in diameter. 36 A series of trials have demonstrated that berberine has several antibacterial mechanisms. Berberine use can lead to DNA damage in bacteria. [37][38][39] The inhibition of FtsZ, the cell division protein, was found to be the main mechanism of action of berberine against bacteria. 40 Studies have shown that berberine can effectively treat infections with drug-resistant bacteria by exerting synergistic effects with different antibiotics. 41 Moreover, studies have demonstrated that berberine can modulate the cytokines in cells and decrease the level of free radicals in the mucous membrane. Moreover, high concentrations of berberine can be absorbed rapidly in the gastric mucosa with long-lasting effects. 42 Berberine has also been found to alleviate antibiotic-related diarrhea by modulating the intestinal microbiome. 43 Furthermore, berberine has been proven to relieve diarrhea by inhibiting the contraction of intestinal smooth muscle cells. 44 A metaanalysis has also demonstrated that berberine-containing therapies will enhance the efficacy and decrease the incidence of adverse effects. 45 To attain high-level clinical evidence, we conducted this large, multicenter, prospective, randomized, controlled, noninferiority study. Regarding symptom improvement, we found that in more than 60% and more than 80% of patients the symptoms significantly improved at 2 and 6 weeks after the initiation of the treatment, respectively.
Although the two regimens did not show any significant statistical difference, berberine and amoxicillin showed relatively higher rates of symptom improvement compared with tetracycline and furazolidone.
Three of five participants in the TF group failed to achieve H. pylori eradication and received the BA regimen, achieving successful eradication 2-3 months later. These findings indicate that BA-based quadruple therapy has the potential to overcome resistance to tetracycline or furazolidone. More importantly, most of our participants (>90%) had used amoxicillin-containing therapy in previous eradication regimens, which may indicate that berberine is capable of overcoming resistance to amoxicillin or even of eradicating H. pylori without the use of other antibiotics.
Previous studies have indicated that the number of treatments and previous therapies may influence the efficacy of rescue eradication therapy. In our trials the previous therapies and numbers of treatment were comparable in both groups.
There were some limitations to our study. First, our centers were all located in northwest China, and most patients were from the area surrounding Shaanxi Province. Whether the berberine plus amoxicillin regimen is available in other countries or other areas in China needs to be confirmed. Second, the dosage of berberine was determined based on our experience; whether higher doses of berberine can improve efficacy requires further research.
Moreover, a series of assessments of the factors affecting the eradication efficacy of the two regimens, such as polymorphisms of CYP2C19 virulence factors are needed. In future studies we will assess the factors that may help us select the best antibiotics and proton pump inhibitors. In addition, we did not complete H. pylori culture, which may result in difficulties in analyzing whether susceptibility and drug resistance of H. pylori strains were balanced between the groups. Therefore, we intend to culture H. pylori strains from participants to analyze the characteristics of H. pylori strains in further trials.

| CONCLUSIONS
In conclusion, to our knowledge, our study is the first to provide an innovative approach to the rescue therapy for H. pylori eradication involving the addition of a traditional Chinese medicine. Based on our results, the efficacies of these two regimens were satisfactory, and berberine plus amoxicillin quadruple therapy was noninferior to tetracycline plus furazolidone quadruple therapy.

CONFLICT OF INTEREST
None.