With the AST of 153 H. pylori strains isolated from patients in Tibet Autonomous Region, China, we found a very high resistance to MTZ (86.2%) and RIF (73.2%), a relatively high resistance to LEV (31.3%) and CLR (21.4%), a moderate resistance to AML (15.6%), and low resistance to TET (0.0%) and FZD (0.6%). In addition, the high prevalence of MDR strains should be highlighted (35.3% of strains were resistant to three, four, and five antibiotics). The results demonstrate a treatment challenge of H. pylori infection due to the high resistance to the commonly used antibiotics.

CLR and MTZ are two of the most commonly used antibiotics for respiratory tract or anaerobic infections. Frequent use of these antibiotics for these infections contributes to H. pylori resistance [19]. Although regionally variable, according to a meta-analysis of 178 studies, the resistance rates of H. pylori strains to them are increasing to more than 15% in nearly the whole world [10]. With regard to resistance in the northern, southeastern, and central regions of China, H. pylori resistance rates to CLR and MTZ were reported to be 19%-45% and 74–89%, respectively [12]. In our current study, the CLR resistance rate of 21.4% and MTZ of 86.2% in the Tibet Autonomous Region were within the range of those in other regions. Moreover, previous studies have shown that female patients carry higher CLR and MTZ resistance rates than male patients due to the higher incidence of treating gynecological diseases with the two antibiotics [20, 21]. Here, although the resistance rate between genders was not statistically significant, higher resistance to CLR and MTZ in female patients than in male patients was still observed, which was in line with the results of previous studies [20, 22]. As CLR resistance critically decreases the efficacy of standard CLR-based triple or quadruple therapy, the present result suggests the avoidance of CLR-containing regimens for H. pylori eradication without prior susceptibility testing in Tibet Autonomous Region, China. Although the eradication rate of H. pylori by MTZ-containing therapy can be improved by increasing its dosage, prolonging its duration or adding bismuth, the increased incidence of adverse side effects such as nausea, vomiting, swirling and rashes needs to be considered [23]. Therefore, it would be better to abandon MTZ-containing regimens as empirical treatment regimens for H. pylori eradication without prior susceptibility testing.

LEV-containing triple or quadruple regimen has been increasingly used for second-line eradication therapy [4]. A previous European multicenter study demonstrated a relatively low rate of LEV resistance of 14% [22]. Nevertheless, higher LEV resistance rates have been reported in Asia (up to 56% in China, 34% in Japan, and 28% in Korea) [6]. In this study, a high rate (31.1%) of LEV resistance was also observed, which surpasses the 25% threshold of resistance rate for choosing LEV as part of empirical or rescue regimens [24], discouraging LEV-containing regimens appropriate for H. pylori eradication without prior susceptibility testing. Of note, a higher rate (40.5%) of LEV resistance in female patients than that in male patients (21.5%) needs special attention, which may also be related to more consumption of LEV because of other respiratory tract, urinary tract or gynecological infectious diseases in female patients [20, 21].

In the current study, the rate of resistance to AML (15.6%) in the Tibet Autonomous Region, China was comparable to that reported in the Eastern Mediterranean region (14%) [10]. What should be noted is that in this study, AML resistance was defined when the MIC was more than the EUCAST recommended cut-off value of 0.125 mg/L. When we applied MIC > 1 mg/L as the AML resistance cut-off value used in a multi-region study [25], 23 strains with MICs of 0.19, 0.25, 0.5 and 1 mg/L were redefined as susceptible, giving an AML resistance rate down to 0.6% in the Tibet region, comparable to the overall AML resistance rate of 3.4% in China [25]. Hence, the AML resistance role is negligible in clinical practice, which means that this antibiotic could be prescribed in most cases.

Similar to AML, the TET and FZD resistance rates were also low at 0% and 0.6%, respectively, which were in accordance with the resistance rates reported in other regions of China [9, 13]. Regimens including these two antimicrobials without prior susceptibility testing are expected to achieve high eradication efficacy in patients there. Unfortunately, they are not generally available in China, affecting their clinical application. Semisynthetic TET derivatives, including minocycline, are easily obtained in the clinic, and their antibacterial activity is higher than that of TET. It is interesting to test the sensitivity of H. pylori to the TET derivatives to determine whether they can be used as alternatives of TET for treating H. pylori infection.

Rifamycins, including RIF, rifabutin, rifaximin and rifapentine, are transcriptional inhibitor antibiotics that suppress bacterial DNA-directed RNA polymerase [26]. Among them, rifabutin plays an important role in salvage treatment of H. pylori infection [5, 27]. Previous studies demonstrated that the resistance rate in H. pylori to RIF was 1.0%, with full agreement with the rifabutin resistance rate (1.0%) when the breakpoint of RIF resistance was > 4 mg/L [26]. Therefore, they considered that there is cross-resistance between RIF and rifabutin in H. pylori and suggested RIF susceptibility testing to be used for screening rifabutin resistance [26]. In this study, we found that the resistance rate of H. pylori against RIF was much higher (73.2%) than the previously reported rate of 1.0% [28, 29]. Even though the breakpoint of RIF resistance > 4 mg/L was used as suggested by Hays et al. [26], resistance to RIF was still high at 32.0%. The higher RIF resistance rate might be because of two reasons. 1) In many previous articles, resistance to RIF was still detected using the disk-diffusion method (despite the MIC value provided by EUCAST) [30, 31], however, E-test was used to detect RIF resistance in our study. Hence the frequency of resistance may be different. 2) The wide use of RIF for treating tuberculosis, a disease with high prevalence in Tibet Autonomous Region, China, leads to a higher resistance rate of H. pylori to RIF [32]. Of note, if cross-resistance between rifabutin and RIF truly exists, the high RIF resistance rate in H. pylori strains in Tibet, China, would indicate the severity of rifabutin resistance. Nevertheless, it is worth noting that the rifabutin MICs of RIF-resistant strains were not known in previous studies [26, 33], and cross-resistance between RIF and rifabutin needs to be further confirmed in RIF-resistant strains isolated from the Tibetan region of China.

The comprehensive information on H. pylori resistance to 7 antibiotics in this study will be very helpful to select the optimal eradication regimens in the Tibet Autonomous Region, China. Nevertheless, two limitations of our study should be noted. First, we did not know whether the patients had undergone eradication therapy prior to this study. Therefore, the primary or secondary H. pylori resistance rate was not confirmed. Second, our findings were based on adult patients in a single hospital. Most of the included patients were from Lhasa city and only few were from other parts of Tibet, so this study may not reflect the general resistance of H. pylori strains in the whole population of the Tibetan region. Future studies are needed to include more patients (with the inclusion of less than 18 years old and from more hospitals) to determine the general antibiotic resistance in H. pylori strains from Tibet Autonomous Region, China.