The potential for development of clinically relevant microbial resistance to rifaximin-α: a narrative review

SUMMARY

Rifaximin-α is a gut-targeted antibiotic indicated for numerous gastrointestinal and liver diseases. Its multifaceted mechanism of action goes beyond direct antimicrobial effects, including alterations in bacterial virulence, cytoprotective effects on host epithelial cells, improvement of impaired intestinal permeability, and reduction of proinflammatory cytokine expression via activation of the pregnane X receptor. Rifaximin-α is virtually non-absorbed, with low systemic drug levels contributing to its excellent safety profile. While there are high concentrations of drug in the colon, low water solubility leads to low colonic drug bioavailability, protecting the gut microbiome. Rifaximin-α appears to be more active in the bile-rich small bowel. Its important biologic effects are largely at sub-inhibitory concentration. Although in vitro testing of clinical isolates from rifaximin recipients has revealed rifaximin-resistant strains in some studies, the risk of emergent rifaximin-α resistance appears to be lower than for many other antibiotics. Rifaximin-α has been used for many years for traveler’s diarrhea with no apparent increase in resistance levels in causative pathogens. Further, rifaximin-α retains its efficacy after long-term and recurrent usage in chronic gastrointestinal disorders. There are numerous reasons why the risk of microbial resistance to rifaximin-α may be lower than that for other agents, including low intestinal bioavailability in the aqueous colon, the mechanisms of action of rifaximin-α not requiring inhibitory concentrations of drug, and the low risk of cross transmission of rifaximin-α resistance between bacterial species. Reported emergence of vancomycin-resistant Enterococcus in liver-disease patients maintained on rifaximin needs to be actively studied. Further studies are required to assess the possible correlation between in vitro resistance and rifaximin-α efficacy.

INTRODUCTION

Rifaximin-α, an orally administered, gut-targeted antibiotic of the rifamycin class, has broad-spectrum activity against Gram-negative and Gram-positive bacteria and is indicated for a number of gastrointestinal and liver diseases, including the treatment of traveler’s diarrhea and irritable bowel syndrome (IBS), and the treatment and prevention of hepatic encephalopathy in end-stage liver disease (1, 2). The agent has been in use for over 30 years and shows an excellent safety profile; in clinical trials, the frequency of adverse events with rifaximin-α was similar to that with placebo (2). The mechanism of action of rifaximin-α in the management of gastrointestinal disorders goes beyond direct antimicrobial effects (1), and this has led to the agent being investigated for numerous infectious and non-infectious conditions affecting the gastrointestinal tract where the drug is given short term (traveler’s diarrhea), chronically (hepatic encephalopathy), and intermittently (diarrhea-predominant IBS).

For all antibiotics, the possibility of development of antimicrobial resistance is a concern. Although the risk of clinically relevant microbial resistance to rifaximin-α appears to be lower than that for many other antibiotics (3), it is important to regularly monitor this risk, particularly in light of indications requiring long-term or repeat administration of the agent. This narrative review provides an overview of the relationship between the pharmacokinetics of rifaximin-α and its safety profile and potential for development of resistance, a summary of the agent’s known and possible mechanisms of action, and a discussion of observed microbial resistance to the drug in isolates from rifaximin recipients, along with an assessment of the clinical relevance of these observations.

DATA SOURCES AND SEARCHES

A PubMed search was conducted on 23 February 2021 for English-language articles. Search terms included rifaximin in combination with (using the Boolean operator “AND”) resistance or clinical resistance. The reference lists of recent articles, including review articles, were also searched. The aim of these searches was to identify articles reporting on the development of microbial resistance to rifaximin in human subjects or patients receiving rifaximin. The author’s knowledge of the subject was also used to identify any articles that the searches may have missed.

MECHANISMS OF ACTION OF RIFAXIMIN-Α

Antimicrobial effects

Rifaximin-α exerts its antimicrobial effects by inhibiting the DNA-dependent RNA polymerase of target bacteria, leading to suppression of bacterial RNA synthesis (4). In clinical trials, rifaximin-α was effective in reducing symptoms of traveler’s diarrhea, even though stool samples showed a low rate of pathogen eradication (5). These were the first studies showing that the major effects of rifaximin were not antimicrobial in the classical sense but that the drug worked through alteration of the biology of the infecting strains, including inhibition of bacterial adherence to epithelial cells, reduction in bacterial motility and virulence gene expression, and alteration of the gut-host inflammatory response, and bacterial metabolic function (6–8).

Rifaximin does not damage the integrity and makeup of the intestinal microbiome

Repeated treatment with rifaximin-α was shown to have only a minimal, transient effect on the composition of the fecal microbial community (9). Although rifaximin-α does not cause any marked long-term changes in fecal microbial composition at the phylum or order level (10), it has been shown to promote the growth of beneficial bacteria, such as bifidobacteria and lactobacilli, and thus aid restoration of mucosal integrity (8). Rifaximin’s antimicrobial effects are greatest in the bile-rich small intestine, which may explain the drug’s value in the treatment of IBS and small intestinal bacterial overgrowth (11). With the large concentrations of rifaximin in the lower gut, why are there not greater changes in the microbiome? The reasons undoubtedly relate to the unique pharmacokinetics of the drug. Rifaximin is essentially insoluble in water and the colonic microbiota live in an aqueous environment. Thus, the microbiota see only small quantities of the drug, the portion that is water soluble, which is sufficient for biologic effects.

In an 18-month, double-blind, placebo-controlled clinical trial, the effect of rifaximin in patients with biopsy-established alcohol-induced liver disease was assessed (12). Rifaximin was well tolerated in this trial. In the per-protocol analysis, an increase in fibrosis occurred in 13 (24%) of the subjects randomized to rifaximin and 23 (43%) of those receiving placebo (P = 0.044). The authors concluded that the reduction in fibrosis was likely due to a reduction in hepatic inflammation by rifaximin.

The important biologic effects of rifaximin occur at sub-inhibitory concentrations

The most important biologic effects of rifaximin appear to be unrelated to the drug’s antibacterial properties and are seen at concentrations below inhibitory concentrations (13–15).

Pretreatment of epithelial cells with rifaximin rendered the cells resistant to bacterial attachment and internalization and reduced mucosal inflammation (13). Not only do sub-inhibitory concentrations of rifaximin have favorable biologic effects but also these low concentrations of drug can lower the viability and virulence of bacteria (14). In treating traveler’s diarrhea, rifaximin leads to rapid recovery without pathogen eradication (5), further supporting the hypothesis that rifaximin prevents pathogen attachment to epithelial cells (13, 15).

In chronic liver disease, sub-inhibitory concentrations of rifaximin likely inhibit urease production and other metabolic products involved with hepatic encephalopathy (16).

Summary of the mechanisms of action of rifaximin

The diverse efficacy profile of rifaximin-α may be explained by the fact that its mechanism of action, although incompletely understood, is multifactorial and does not depend upon inhibition of bacterial growth (Fig. 1). Rifaximin-α exerts favorable effects on host epithelial cells and in altering the virulence of pathogens (13, 15). In an in vitro study, while the reduction of bacterial adhesion was attributed primarily to the effects of rifaximin on bacteria (Escherichia coli), the decrease in bacterial internalization into cells was attributed to the drug’s effects on the epithelial cells (7, 13). Another in vitro study showed that pre-treatment of epithelial cells with rifaximin altered the cell physiology in a variety of ways to confer cytoprotection, including down-regulating a marker for apoptosis (suggesting protection from bacterial-induced apoptosis), and beneficial alterations in the expression of proteins related to cellular integrity and the cytoskeleton (17). In patients with decompensated cirrhosis, 4 weeks of rifaximin treatment significantly decreased serum markers for intestinal permeability, suggesting that rifaximin may improve impaired intestinal permeability in these patients (18). In the same trial, rifaximin reduced serum endotoxin and ammonia levels and improved cognitive performance. In a mouse model of colitis, rifaximin treatment reduced the number of bacteria translocating across the epithelial barrier into the mesenteric lymph nodes (19). This finding is significant for inflammatory bowel disease (IBD), as it is thought that translocation of intestinal bacteria into these normally sterile lymph nodes may contribute to the pathophysiology of the disease.

Fig 1.

Putative mechanisms of rifaximin-α on the gut and intestinal bacteria (see text for references). SIBO, small intestinal bacterial overgrowth.

Rifaximin-α is an activator of the pregnane X receptor (PXR); this receptor senses toxic substances and is involved in detoxification (20). Activation of PXR by rifaximin-α can regulate the inflammatory response, including reducing the expression of proinflammatory cytokines (21). This mechanism may explain some of the effectiveness of rifaximin-α in patients with IBD, as many of these patients have abnormally low PXR levels (19).

CHEMISTRY

Figure 2 provides the chemical structure of rifaximin-α. The pyridoimidazole ring is responsible for low water solubility and the limited absorption of this molecule. The formula of rifaximin-α is C₄₃H₅₁N₃O₁₁, and it has a molecular weight of 785.9. Rifaximin [4-deoxy-4′-methylpyrido(1′,2′−1,2)imidazo-(5,4 c)-rifamycin] is derived from rifampin. Rifaximin-α is soluble in methanol, chloroform, acetone, and ethyl acetate. While practically insoluble in water, rifaximin-α is soluble in bile salts (22).

Fig 2.

Chemical structure of the rifaximin molecule. The pyridoimidazole ring, which makes the molecule poorly absorbed, is circled.

PHARMACOKINETICS AND PHARMACODYNAMICS

Two studies provide pharmacokinetic data for two different doses of rifaximin-α. In the first study, 400 mg of rifaximin-α was given to 24 subjects. The mean (±standard error of the mean) highest concentration in blood (C_max) was 2.29 ± 0.28 ng/mL; the time to C_max (T_max) was 1.78 ± 0.46 h; the area under the curve (AUC) from time 0 to the last measurable concentration (AUC_0–t) was 6.34 ± 0.99 ng/mL × h; total drug exposure across time (AUC_0–∞) was 9.80 ± 1.23 ng/mL × h; half-life (T_1/2) was 2.8 ± 0.4 h; and the amount of drug in the urine in 48 h was 45.27 ± 4.61 µg (23).

In the second study, when 550 mg of rifaximin-α was given orally twice a day to 12 healthy subjects, the mean (±standard deviation) C_max was 3.41 ± 1.62 ng/mL; AUC over the dosing interval was 12.3 ± 4.76 ng/mL × h; oral clearance was 863 ± 364 L/min; T_1/2 was 4.14 ± 3.3 h; and the median T_max was 0.76 (range 0.5–4) h (24).

A key feature of rifaximin-α is that it is virtually non-absorbed: it is likely that the resulting low systemic bioavailability accounts for the excellent safety profile of the drug and minimizes the risk of acquisition of extraintestinal antimicrobial resistance, while the gastrointestinal bioavailability is sufficient to provide clinical efficacy (25). Furthermore, the bioavailability of rifaximin-α is higher in the small intestine than in the colon, as the drug has very low solubility in aqueous environments, such as the colon, and higher solubility in the presence of bile acids, as found in the small intestine (11, 22).

It is also important to note that, well after it was approved for use, it was discovered that rifaximin shows crystal polymorphism, with several polymorphs documented. The branded, marketed rifaximin contains only bioactive rifaximin-α. Studies in healthy volunteers have shown that rifaximin-α has lower systemic bioavailability than a generic (23) or amorphous rifaximin (25). It is therefore important that we do not treat all forms of rifaximin as equal, as greater systemic bioavailability has the potential to result in reduced efficacy in gastrointestinal disorders and an increased risk of adverse events and systemic antimicrobial resistance (25).

Minimum inhibitory concentration breakpoints for rifaximin

The low systemic bioavailability of rifaximin-α makes determination of organism-specific antimicrobial breakpoints difficult (26). While the drug is virtually non-absorbed and rifaximin reaches very high concentrations in the gut, as indicated by a study in 39 adults with traveler’s diarrhea in which fecal levels of rifaximin-α reached 4,000–8,000 µg/g after 3 days of treatment (800 mg/day) (27), most of the drug is not bioavailable because of low water solubility in the aqueous environment of the colon. Traditional breakpoints that are based on plasma drug concentrations are not appropriate for rifaximin (1). When minimum inhibitory concentration (MIC) breakpoints are estimated for rifaximin, they are given in the range of ≥32 to ≤64 µg/mL for susceptible organisms (28, 29). These breakpoints have been suggested based on in vitro studies of susceptibility for various strains of bacterial pathogens (30–33) and breakpoint data for related rifampin.

Since rifaximin is not used to eradicate pathogens and the biologic activity of the drug is not dependent upon inhibitory drug concentrations, breakpoints are of little clinical significance for rifaximin.

MECHANISMS OF RESISTANCE TO RIFAXIMIN-α AND CLINICAL RELEVANCE

General

The most common mechanism for the development of resistance to rifaximin-α is a chromosomal one-step alteration in the gene encoding for bacterial DNA-dependent RNA polymerase (rpoB gene), resulting in a resistant sub-optimal enzyme (31). Antibiotic resistance induced by a chromosomal mutation is transferred directly to the bacteria progeny during DNA replication. Rifaximin, like rifampin, can select for resistant bacterial strains in the gut but at low frequency. Spontaneous rifaximin-resistant clones appear with a frequency of 2.6 × 10⁷ (14). In this study, rifaximin-α lowered the viability and virulence of bacteria that became resistance. The mechanism of chromosomal resistance seen in rifamycins differs from plasmid-mediated resistance that is easily acquired by susceptible bacteria after treatment with aminoglycosides, sulfonamides, and macrolides and can be rapidly and widely spread both within and among bacteria species (34). Thus, there is minimal risk of cross transmission of rifaximin-α resistance between bacteria species.

In a study of E. coli isolated from the ilium of patients with IBD, 2 of 12 strains resistant to rifaximin demonstrated Phe-Arg-b-naphthylamide-inhibitable efflux pumps responsible for resistance (35), a finding also reported by another group (36).

Resistance acquisition during therapy with rifaximin

Rifaximin-α has been in clinical use for decades and has been the subject of numerous studies in various therapeutic areas. Results of selected studies examining microbial resistance in human rifaximin-α recipients are presented in Table 1. In an early study in healthy volunteers, microbial resistance to rifaximin-α was not observed at baseline but was observed in 30%–90% of all strains isolated after 5 days of treatment (37). The resistant strains disappeared rapidly after treatment was discontinued.

TABLE 1.

Trials investigating the development of microbial resistance to rifaximin-α in human subjects or patients receiving rifaximin-α^a

Study	Population and treatment	Microbial sampling	Results
(37)	Healthy volunteers, rifaximin 800 mg/day for 5 days (N = 10)	Stool samples collected before treatment and weekly for 16 weeks	Proportions of Enterobacteriaceae, enterococci, Bacteroides spp., Clostridium spp., and anaerobic cocci were reduced from baseline to week 1 but returned to baseline values 1–2 weeks after end of treatment. Microbial resistance was not observed at baseline but was observed in 30%–90% of all strains isolated after treatment. Percentages of resistant strains declined rapidly after treatment discontinuation, and no resistant strains were detected 3 months after treatment.
(38)	Patients with UC, rifaximin 1,800 mg/day for 3× 10-day cycles (washout 25 days between each cycle) (N = 12)	Stool samples collected before treatment (day 0), end of treatment (days 10, 45, and 80), and end of washout (days 37, 70, and 105)	Proportions of enterococci, coliforms, lactobacilli, bifidobacteria, Bacteroides spp., and Clostridium perfringens varied throughout the study, but at the end of the study, they were not significantly different from pre-treatment. Although susceptible to rifaximin in vitro, the proportions of test strains did not significantly alter during therapy and all returned to pre-treatment values after treatment.
(39)	Patients with TD, (i) rifaximin 600 mg/day, (ii) rifaximin 1,200 mg/day, or (iii) placebo for 3 days; patients with both pre- and post-treatment samples available	Stool samples collected on days 0, 3, and 5.	Stools plated on media containing rifaximin or rifampin days 3 and 5 (2 days after stopping treatment) did not yield an increase in resistant E. coli compared to pre-treatment baseline (N = 27). Susceptibility of enterococci grown on days 0 and 3 showed similar susceptibilities to rifaximin (N = 71).
(40)	US adults traveling to Guadalajara, Mexico, randomized to TD prophylaxis with rifaximin 200, 400, or 600 mg/day, or placebo for 2 weeks (N = 210)	Stool samples from all participants on days 7 and 14; additional samples at the time of diarrhea, where applicable	Median log counts of fecal coliform bacteria were similar on days 7 and 14 in rifaximin-treated subjects. Only for the group receiving 600 mg/day on day 7 was the median log value of coliform bacteria grown on medium containing rifaximin higher than placebo (P = 0.01). The rifaximin MIC50 and MIC90 for coliforms and enterococci showed a value one dilution higher than the isolates obtained from the placebo.
(35)	E. coli isolated from ileal mucosa of 50 patients with IBD and 8 healthy controls were studied for susceptibility to rifaximin.	Rifaximin resistance (MIC >1,024 mg/L) was present in 12/48 of ileal E. coli from patients with IBD and none from the controls.	Rifaximin resistance correlated with prior exposure to rifaximin but not presence of ileal inflammation. Mutations in rpoB were identified in 10 of 12 resistant strains compared with 9 of 50 of susceptible strains. The efflux pump inhibitor PAβN lowered the MIC of 9 of 12 resistant strains.
(41)	Patients with IBS-D received rifaximin treatment in an open-label phase, and those responding entered a double-blind, placebo-controlled phase. The rifaximin dose for both phases was 550 mg three times daily for 2 weeks.	A total of 1,429 bacterial and yeast isolates were obtained from stools. The most abundant strains of the families were Bacteroidaceae, Enterobacteriaceae, Enterococcaceae, and Staphylococcaceae.	Although strains within all families showed an increase in resistance to rifaximin during treatment, many strains remained susceptible to both rifaximin and rifampin during the studies. The MIC50 and the MIC90 did not change for any of the organisms compared with baseline values. In the study, there was no evidence of long-term antimicrobial resistance with repeated courses of rifaximin for the treatment of IBS-D.
(42)	In a trial where subjects with IBS-D received three 2-week courses of rifaximin, isolates were cultured from skin; swabs of the perianal area, nostrils, forearms, and palms of hands of subjects on five occasions	A total of 1,381 isolates of Staphylococcus were obtained. MIC50 and MIC90 of the placebo and rifaximin groups were low (≤0.06) from baseline through end of study. Transient increases in MIC were seen in the rifaximin-treated subjects but returned to baseline at end of study.	Short-term (2 weeks) administration of rifaximin (1,650 mg/day) for up to three courses did not lead to clinically significant or persistent resistance to rifaximin or rifampin.
(43)	Retrospective chart review of patients with cirrhosis (N = 388); 127 of 388 had diarrhea with stool sample; 46 of 127 had CDI; 14 of 46 were receiving rifaximin for HE prophylaxis (median 69 days treatment prior to CDI).	C. difficile isolates with MIC >32 mg/L for rifaximin were considered resistant.	In patients with CDI (n = 46), rifaximin-resistant C. difficile strains were detected in 34.1% overall and in 84.6% of those who were receiving rifaximin. Infection by ribotype 001 was significantly associated with rifaximin resistance.

^a CDI, Clostridium difficile infection; HE, hepatic encephalopathy; IBD, inflammatory bowel disease; IBS-D, diarrhea-predominant irritable bowel syndrome; MIC, minimum inhibitory concentration (mg/L); MIC50, MIC required to inhibit the growth of 50% of organisms; MIC90, MIC required to inhibit the growth of 90% of organisms; PAβN, Phe-Arg-β-naphthylamide; TD, traveler’s diarrhea; UC, ulcerative colitis.

When patients with ulcerative colitis received repeat rifaximin-α treatment (3 × 10-day courses, separated by 25-day washout periods), there were minimal changes from study beginning to end in the mean rifaximin concentration at which spontaneous rifaximin resistant clones were isolated; interestingly, the largest increase was for the beneficial bifidobacteria (38).

In patients with traveler’s diarrhea who received rifaximin or placebo for 3 days, there were no significant changes in the numbers of rifaximin-α-resistant coliforms from pre- to post-treatment, and the MICs for enterococci were very similar at baseline and after treatment, regardless of treatment group (39). In another trial in which adults traveling to Mexico received 2 weeks’ treatment of prophylactic rifaximin-α or placebo, pathogens isolated from patients who developed diarrhea had similar MIC values in the rifaximin-α and placebo groups (40). In isolates from asymptomatic patients, susceptibility to rifaximin-α was slightly lower in the rifaximin-α group than in the placebo group.

Two sub-studies used samples from patients enrolled in the TARGET 3 study, in which patients with diarrhea-predominant IBS (IBS-D) received treatment with rifaximin in the first open-label phase, and then repeated treatment with rifaximin or placebo in the double-blind phase (41, 42). In one sub-study, Staphylococcus isolates from skin swabs (peri-anus, nostrils, forearms, and palms) of patients receiving rifaximin had a somewhat reduced susceptibility to rifaximin after treatment, but the MIC values were very low by the end of the study (42). Similarly, in the other sub-study, MIC₉₀ values for Staphylococcaceae increased after treatment but had dropped to the low baseline levels by the end of the study (41). There were minimal changes in the MIC₉₀ values for Enterococcaceae throughout the study. In contrast, MIC₉₀ values for Bacteroidaceae and Enterobacteriaceae increased after treatment with rifaximin-α in the open-label phase of the trial and remained at this higher value in both the rifaximin and placebo groups throughout the double-blind phase of the trial.

Of some concern is evidence showing the emergence of rifampin resistance in staphylococcal isolates from patients with cirrhosis taking rifaximin 400 mg three times daily to prevent hepatic encephalopathy (44). Between 1 and 7 weeks after initiating rifaximin treatment, testing of isolates from skin swabs (hand and perianal regions) showed that 11 of 25 (44.0%) patients had developed rifampin-resistant staphylococcal strains, with a median MIC of 32.0 (interquartile range 4.0–32.0) μg/mL (44). While no new resistant strains were identified from samples taken 8–16 weeks after initiating rifaximin treatment, 6 of the 11 patients continued to show rifampin-resistant staphylococcal isolates.

In an analysis of E. coli isolates from 50 participants (IBD, symptomatic non-IBD, or healthy), rifaximin-resistant strains were found in 11 of the 13 participants who had received rifaximin and in none of those who were rifaximin naïve; the MIC values for all resistant strains were >1,024 mg/L, but the presence of rifaximin-resistance did not correlate with ileal inflammation (35). This again suggests that poorly absorbed rifamycin has important biologic effects beyond strict pathogen inhibition.

Finally, in a retrospective chart review of 388 patients with cirrhosis, of whom 46 had confirmed Clostridium difficile infection, rifaximin-resistant C. difficile strains were detected in 34.1% overall and in 84.6% of those who had received rifaximin (43). Resistance of strains of C. difficile to rifaximin appears to vary by geographic region: examination of C. difficile clinical isolates from patients with C. difficile infection found a higher frequency of resistant isolates in Italy than Canada (18.8% vs 2.4%) and a higher level of resistance (MIC₉₀ of >16 µg/mL vs 0.0008 µg/mL) (45). Indirect evidence from this study suggested that higher exposure at a population level to rifaximin may have resulted in a positive selection pressure for rifamycin resistance in Italy, because at the time the study was conducted, rifamycin antibiotics had been in use in Italy for decades, whereas rifamycin was less commonly used in Canada, and unlike in Canada, rifaximin was already licensed for multiple indications (hepatic encephalopathy, diarrhea, traveler’s diarrhea, and prophylaxis prior to gastrointestinal surgery) (45). However, these results should be interpreted with caution since samples were from only a single center in each country, and newer studies in current clinical settings across a range of geographic locations would be required to confirm these findings. Given the limited data on rifaximin resistance, it is interesting to note that this latter study demonstrated that susceptibility to rifampin correlated to susceptibility to rifaximin (at least when E-test or agar dilution testing is used), suggesting data on rifampicin resistance may predict rifaximin resistance (45).

With, perhaps, the exception of the traveler’s diarrhea prophylaxis trial, in which pathogens were isolated from those in whom prophylaxis failed, none of these trials investigated the clinical relevance of the findings, that is, the correlation between efficacy or treatment failure and observed resistance. As discussed earlier, the rifaximin-α concentration in the gastrointestinal tract after oral administration is likely to be extremely high, while intracolonic bioavailability would be lower, making it difficult to interpret the significance of in vitro susceptibility testing (31).

There are several reasons to suggest that there may be a low risk of development of clinically relevant microbial resistance to rifaximin-α (Table 2) (10, 31, 34, 46–49). These include that the mechanisms of action of the drug go beyond direct antibacterial effects (thus possibly making in vitro resistance less of a concern); that rifaximin-α resistant microbes may be less “fit” and, therefore, other gut bacteria may have a competitive advantage; and that the drug shows low water solubility, meaning concentrations reached in aqueous environments are likely to be too low to encourage resistance.

TABLE 2.

Suggested reasons for the low risk of development of clinically relevant microbial resistance to rifaximin-α^a

Mechanism	Comment
Rifaximin-α is essentially non-absorbed. Rifaximin-α shows low water solubility (10), minimizing drug concentrations in the aqueous colon.	Systemic resistance, outside the gut, is unlikely to occur. Bacterial responses to antimicrobial agents are concentration dependent (46). Sub-inhibitory concentrations of rifaximin alter organisms’ virulence while preventing widespread major changes in the microbiome and persistence of resistant strains.
One-step chromosomal resistance to rifamycins involving rpoB gene is not associated with mobilization to other strains (31, 34).	Acquisition of rifaximin-α resistance is transient, disappearing when the drug is stopped, and resistance is not mediated by plasmid/transposon factors, which prevents spread in the gut to other bacterial strains.
Metabolic modifications of antibiotic-resistant mutants minimize persistence.	Resistant mutants with their membrane saturated fatty acids (47) lack “fitness” and are not given selective advantage in the face of continued rifaximin-α therapy.
Mycobacterium tuberculosis shows low potential for development of resistance to rifaximin-α (48).	Rifaximin does not select for resistant strains of M. tuberculosis in artificial media (48) or in experimental animals (49). The non-systemic nature of the drug is a deterrent to emergence of resistance for this pathogen.

^a rpoB, bacterial DNA-dependent RNA polymerase.

Although rifaximin-α has been used for many years for the treatment or prophylaxis of traveler’s diarrhea, the levels of rifaximin-α resistance in pathogens causing traveler’s diarrhea do not appear to have increased over the years. In two studies conducted approximately 10 years apart in the same geographic regions by the same research team, 1997 (30) vs 2006–2008 (32), there were no significant increases in the rifaximin MIC values of isolates from patients with traveler’s diarrhea. In contrast, the MIC values for other commonly used antibiotics, including fluoroquinolones and azithromycin, were significantly higher in the second study than the first (30, 32). Similarly, a study of 39 enteroaggregative and 43 enterotoxigenic E. coli strains (isolated 2011–2017) from patients with traveler’s diarrhea found that no strains had a rifaximin MIC of >32 µg/mL (50).

Furthermore, clinical trials have shown that rifaximin-α can be used for long-term maintenance or repeat treatment and retain efficacy. For example, in one trial in which patients with hepatic encephalopathy received rifaximin-α for over 2 years, efficacy at reducing hospitalization was retained (51). A 2019 systematic review found that long-term (≥6 months) rifaximin added to lactulose was significantly more effective than lactulose monotherapy at reducing hepatic encephalopathy recurrence and related hospitalizations (52). In a retrospective study in 346 patients with symptomatic uncomplicated diverticular disease who received rifaximin for 7 days per month, the 8-year follow up showed rifaximin was effective at relieving symptoms of pain and bloating and reduced the risk of complications (53). In a 6-month study, rifaximin-α was significantly more effective than norfloxacin at reducing the recurrence of spontaneous bacterial peritonitis in patients with liver cirrhosis and ascites (54). Finally, in IBS-D patients who initially responded to rifaximin but then relapsed during treatment-free follow up, repeat treatment with rifaximin was significantly more effective than placebo at reducing abdominal pain, preventing disease recurrence, and providing sustained symptomatic relief (55).

Patients with end-stage liver disease frequently harbor antibacterial-resistant microbes including Enterococcus faecalis or the more antibiotic-resistant Enterococcus faecium (56). Future research should be performed to confirm the observation of an association between rifaximin use and the emergence of rifamycin-resistant E. faecium that are resistant to daptomycin (Turner et al., Rifaximin prophylaxis causes resistance to the last-resort antibiotic daptomycin. medRxiv, posted 5 March 2023; https://doi.org/10.1101/2023.03.01.23286614), a critically important antibiotic to treat these infections. There are additional studies of vancomycin-resistant Enterococcus (VRE) in liver-disease patients given rifaximin. In the first study, administration of antibiotics including rifaximin was associated with an increased rate of pre-transplant VRE acquisition in patients undergoing liver transplantation (57). Similarly, in a second study of patients with cirrhosis, receipt of rifaximin was associated with acquisition of VRE infection (58). In a third study, rifaximin provided a protective effect against early post-transplant infection by multidrug-resistant bacteria including VRE (59), and a separate study also found that rifaximin in the pre-liver transplant period was not associated with increased risk of bacterial or fungal infections when administered early post-liver transplantation (60). In healthy travelers with diarrhea taking short-course rifaximin, VRE acquisition was not increased by administration of rifaximin (39). Acquisition of VRE and daptomycin resistance will have to be watched closely and must be considered in the assessment of the value of rifaximin given long term to prevent hepatic encephalopathy.

Cyclosporine administered after transplantation has been shown to increase the administered dose of rifaximin (61). While cyclosporine increases the bioavailability of rifaximin, the poorly absorbed drug was used successfully in one study to reduce graft injury in liver transplantation (62). Rifaximin provided a cytoprotective effect against liver injury by suppressing graft-associated inflammation. Rifaximin has little use in patients after liver transplantation, where systemically absorbed antibiotics are needed to treat infections (63) unless patients experience rejection and are put back at risk of encephalopathy.

While antibacterial resistance has been reported in patients with end-stage liver disease, given long-term rifaximin, clinical resistance, where treatment with rifaximin becomes ineffective over time, has not been identified. This provides indirect evidence that the beneficial effect of rifaximin in preventing hepatic encephalopathy is either not related to antimicrobial effects of the drug or that sub-inhibitory effects of the drug do not require organism susceptibility, which has been seen for rifaximin-induced lowering of viability and virulence in bacteria even when resistant to the drug (14).

Performing rifaximin in vitro susceptibility of bacterial strains

Rifaximin susceptibility of a bacterial strain can be performed by standard agar dilution testing according to the methods of the National Committee for Clinical Laboratory Standards. The critical part of the assay is making a stock solution of rifaximin using acetone to put the drug into solution (31). Additionally, E-test strips of rifampin have been used to test for rifaximin susceptibility, as discussed above (45).

CLINICAL APPLICATION OF RIFAXIMIN

In the US and many other countries, rifaximin is approved for use in three medical conditions (the references given after the indication represent the clinical trials used in support of a claim for licensure):

• Treatment of traveler’s diarrhea caused by non-invasive strains of E. coli in individuals 12 years of age or older (5, 64).

• Reduction of the risk of overt hepatic encephalopathy recurrence in adults (51, 65).

• Treatment of diarrhea-predominant IBS (55, 66).

Rifaximin is also used in a number of non-approved off-label indications: small bowel bacterial overgrowth (67), diverticulitis (68), and IBD and pouchitis (69, 70).

While rifaximin is approved in the US for use in children 12 years of age and older for traveler’s diarrhea (the only approved use for children in the US), in many countries, it is used off-label in children for the same conditions for which the drug is licensed in adults (71, 72).

CONCLUSIONS AND FUTURE RESEARCH QUESTIONS

Rifaximin-α is an antibiotic that has been used for many years for varied indications and has a multifactorial mechanism of action that goes beyond direct antimicrobial effects. As with all antibiotics, the potential development of microbial resistance to rifaximin-α is a concern, particularly as the drug is being used or investigated for indications requiring long-term maintenance or repeat treatments.

The risk of emergent rifaximin-α resistance appears to be lower than that for many other antibiotics. Indeed, rifaximin-α has been used for many years for the treatment or prophylaxis of traveler’s diarrhea, and the levels of rifaximin-α resistance in pathogens causing traveler’s diarrhea do not appear to have increased. Some in vitro studies of clinical isolates have shown no resistance to rifaximin after treatment or resistance that disappears rapidly after treatment ends. However, in some studies, in vitro testing of clinical isolates from rifaximin recipients has revealed rifaximin-resistant strains. The clinical relevance of this resistance is not clear, as the studies were not designed to assess the correlation of resistance and efficacy. Because of rifaximin-α’s unique pharmacokinetics and mechanisms of action, it appears that the development of microbial resistance to rifaximin-α is not as clinically relevant as it is for most other antimicrobial agents. Rifaximin-α may actually inhibit the development of resistance in the intestinal tract. Sub-inhibitory concentrations of rifaximin-α have been shown to block plasmid transfer between donor and recipient bacteria, even when the bacteria are resistant to the drug (14). The biggest concern for resistance development is in patients with end-stage liver disease who normally harbor antibiotic-resistant bacteria and resistance genes (73). Studies assessing the correlation between in vitro resistance and efficacy are required.

Biography

Herbert L. DuPont is currently Professor of Infectious Diseases, University of Texas – School of Public Health; the Mary W. Kelsey Chair in the Medical Sciences, University of Texas McGovern Medical School; Clinical Professor, Baylor College of Medicine; and Adjunct Professor, MD Anderson Cancer Center. Dr. DuPont served as president of the Infectious Diseases Society of America, the National Foundation for Infectious Diseases, the American Clinical and Climatological Association, and was elected to Fellowship, American Academy of Microbiology. Honors received by Dr. DuPont include: the Maxwell Finland Award for Scientific Achievement from the National Foundation for Infectious Diseases; Mastership in the American College of Physicians; the Alexander Fleming Award for Lifetime Achievements in Infectious Diseases from the Infectious Diseases Society of America; the University of Texas System Regents’ Outstanding Teaching Award; and given the title “Distinguished Teaching Professor”. He has received honorary doctorates from the University of Zurich and Ohio Wesleyan University.