In Vitro Activities of Combinations of Rifampin with Other Antimicrobials against Multidrug-Resistant Acinetobacter baumannii

Abstract

The antimicrobial treatment of multidrug-resistant (MDR) Acinetobacter baumannii infections has become a great challenge for medical staff all over the world. Increasing numbers of MDR A. baumannii infections have been identified and reported, but effective clinical treatments for them are decreasing. The objective of this study was to investigate the in vitro activities of combinations of rifampin (an established antimicrobial) and other antimicrobials, including biapenem, colistin, and tigecycline, against 73 clinical isolates of MDR A. baumannii. In total, 73 clinical isolates of MDR A. baumannii were collected from two A-level general hospitals in Beijing, and the MICs of rifampin, biapenem, colistin, and tigecycline were determined. The checkerboard method was used to determine the fractional inhibitory concentration indices (FICIs), that is, whether the combinations acted synergistically against these isolates. The MIC₅₀, MIC₉₀, and MIC_range of rifampin combined with biapenem, colistin, and tigecycline against the isolates were clearly lower than those for four antimicrobials (rifampin, biapenem, colistin, and tigecycline) that were used alone. Combinations of rifampin with biapenem, colistin, and tigecycline individually demonstrated the following interactions: synergistic interactions (FICI ≤ 0.5) for 31.51%, 34.25%, and 31.51% of the isolates, partially synergistic interactions (0.5 < FICI < 1) for 49.31%, 43.83%, and 47.94% of the isolates, and additive interactions (FICI = 1) for 19.18%, 21.92%, and 20.55% of the isolates, respectively. There were no indifferent (1 < FICI < 4) or antagonistic (FICI ≥ 4) interactions. Therefore, combinations of rifampin with biapenem, colistin, or tigecycline may be future therapeutic alternatives for the treatment of MDR A. baumannii infections.

INTRODUCTION

Acinetobacter baumannii is the most important pathogenic bacterium of the 21st century. Multidrug-resistant (MDR) isolates can be rapidly acquired and disseminated worldwide (1). The development of multidrug, extensive, and complete resistance in A. baumannii and the appearance of pandemic strains have become significant challenges throughout the world, which raises a lot of discussion in the research field. A. baumannii is the most well-known “superbacterium” in China; it can cause acquired pneumonia, blood infections, abdominal infections, central nervous system infections, urinary system infections, and skin and soft tissue infections (2). Moreover, it is a conditioned pathogen of hospital-acquired and respirator-related pneumonia, and it is easily implanted into the skin, conjunctiva, oral cavity, respiratory tract, gastrointestinal tract, and urogenital tract (3). A pandemic of imipenem-resistant A. baumannii outbreaks in hospitals has been reported (4). A recent study suggested that MDR A. baumannii isolates are susceptible to tigecycline and colistin (5). Unfortunately, colistin is not yet available in China. According to reports, a combination of sulbactam (a β-benzosultam inhibitor) with another antimicrobial is an effective treatment for MDR A. baumannii infection (6) that is extensively or entirely resistant, but more clinical research is required to verify this. In 2012, the Chinese CHINET, which monitors data on drug resistance, reported an increase in the rate of carbapenem resistance of A. baumannii to 60%.

Therefore, MDR A. baumannii infections continue to present a great challenge. Rifampin is an established and commonly used drug with antimicrobial properties against Gram-negative and Gram-positive bacteria. The efficacy of rifampin against MDR A. baumannii is noteworthy, but rifampin should not be used alone because a study showed that in treating experimental murine pneumonia caused by A. baumannii, monotherapy with rifampin leads to tolerance after 24 h of use (7). In this case, it is necessary to add another antimicrobial to rifampin in order to prevent the development of resistance. Colistin is another established antimicrobial drug which was denounced in the 1960s for its contribution to intoxication-induced renal damage and has received renewed attention for its activity against A. baumannii (7). Tigecycline is the last defense against A. baumannii and also shows good in vitro bacteriostatic activity against imipenem-resistant A. baumannii (5). Colistin and tigecycline are very effective and are a last defense against A. baumannii. Biapenem is a novel broad-spectrum carbapenem recently introduced for clinical use and has been widely used to treat bacterial pneumonia without serious adverse events, such as nephrotoxicity. It can either reduce the clinical symptoms or eradicate the causative organisms (8). Therefore, it is hypothesized that its inclusion in a combination therapy might be effective against MDR A. baumannii strains.

MATERIALS AND METHODS

Bacterial isolates.

In this study, 73 MDR isolates of A. baumannii collected from two A-level general hospitals in Beijing were evaluated (52 isolates from the Chinese PLA General Hospital and 21 isolates from Beijing Union Medical College Hospital). All the strains were obtained from April to June 2010. The clinical isolates were identified by Vistem (bioMérieux, Marcy-l'Étoile, France). All of them were MDR A. baumannii strains, defined as susceptible to colistin but nonsusceptible to at least one agent in more than three antimicrobial categories (aminoglycosides, antipseudomonal carbapenems, antipseudomonal penicillins plus β-quinolones, β-lactams, extended-spectrum cephalosporins, folate pathway inhibitors, penicillins plus β-lactamase inhibitors, and tetracyclines) (9). Escherichia coli strain ATCC 25922 was used as the quality control strain in each batch of tests.

Antimicrobials.

The four antibacterial agents examined in this study were rifampin, biapenem, colistin, and tigecycline. Rifampin was obtained from the National Institute for the Control of Pharmaceutical and Biological Products (Beijing, China). Colistin was purchased from Sigma-Aldrich (Munich, Germany). Biapenem was purchased from Zheng Da Tian Qing Pharmaceutical Co. (Jiangsu, China), and tigecycline was purchased from the Pfizer pharmaceuticals group (Wyeth, Princeton, NJ, USA). The antimicrobial powders were used to prepare stock solutions at concentrations of 1,024 μg/ml, as recommended by the Clinical and Laboratory Standards Institute (CLSI) (10), and the tigecycline stock solution was prepared according to the manufacturer's recommendations and stored until use. The solvents were distilled water for colistin, biapenem, and tigecycline and methanol (maximum concentration, 640 μg/ml) in distilled water for rifampin.

MIC determination.

Initially, the MICs of any of the four antimicrobials were tested against 2-fold serial dilutions of all the isolates using the broth microdilution method per CLSI standards. The following MICs were determined as the susceptibility breakpoints of the tested antimicrobials: ≤2 μg/ml for colistin (per CLSI standards) (10), ≤2 μg/ml for rifampin (11), ≤4 μg/ml for biapenem (based on previous studies) (12), and ≤1 or 2 μg/ml for tigecycline (per the European Committee on Antimicrobial Susceptibility Testing or the U.S. Food and Drug Administration, respectively) (13, 37).

Synergy testing of rifampin combinations with the checkerboard method.

The synergy tests were performed in 96-well microdilution plates containing rifampin and one of the three antimicrobials (biapenem, colistin, or tigecycline) in 2-fold dilutions dispensed in a checkerboard configuration. Rifampin (0.5 to 32 μg/ml concentration) was combined with biapenem (0.25 to 128 μg/ml), colistin (0.03125 to 16 μg/ml), or tigecycline (0.03125 to 16 μg/ml). The range of concentrations was determined per MIC of each antimicrobial for each test strain in the preliminary susceptibility tests. The bacterial suspensions were prepared by adding colonies from agar plates into Mueller-Hinton II broth (BD BBL, Sparks, MD) and culturing them overnight to a 0.5 McFarland standard. The cultures were diluted to a final concentration of approximately 1.5 × 10⁵ CFU/ml with a multipoint inoculator. Two antimicrobial agents were prepared in a total volume of 100 μl (a fixed volume of up to 50 μl for each antimicrobial agent) and added to each well with 100 μl of bacterial suspension. The microplates were incubated aerobically at 35°C for 18 to 24 h. Escherichia coli strain ATCC 25922 was used as the standard quality control strain in each test batch. The interaction between rifampin and the other antimicrobials (biapenem, colistin, or tigecycline) was determined based on the calculated fractional inhibitory concentration index (FICI) (14). The FICI is the sum of the fractional inhibitory concentration (FIC) of drug A plus drug B. The FIC of each drug was calculated as a ratio of the MIC of drug A (or B) when used in the combination and the MIC of drug A (or B) when used alone, according to the following formula: FICI = FICA + FICB (i.e., [MIC of drug A in combination/MIC of drug A alone] + [MIC of drug B in combination/MIC of drug B alone]).

The FICI results for each combination to each test isolate were determined as follows (14–16): synergism, FICI ≤ 0.5; partial synergism, 0.5 < FICI < 1; additivity, FICI = 1; indifference, 1 < FICI < 4; and antagonism, FICI ≥ 4. The results are expressed as percentages and cumulative inhibition ratios (CIRs) of isolates with synergism, partial synergism, additivity, indifference, and antagonism.

RESULTS

MICs for rifampin combined with other antimicrobials against MDR A. baumannii.

Before combination, the MIC₅₀ and MIC₉₀ values were 4 and 8 μg/ml for rifampin, 32 and 64 μg/ml for biapenem, 0.5 and 1 μg/ml for colistin, and 1 and 2 μg/ml for tigecycline. When combined, the MIC₅₀ and MIC₉₀ values were 1 and 2 μg/ml for rifampin, 8 and 16 μg/ml for biapenem, 0.125 and 0.25 μg/ml for colistin, and 0.25 and 0.5 μg/ml for tigecycline.

When the antimicrobials were used alone, the MIC_range values and susceptibility rates of MDR A. baumannii for nontraditional antibiotics were 1 to 16 μg/ml and 30% for rifampin, 1 to 64 μg/ml and 5.5% for biapenem, 0.125 to 1 μg/ml and 100% for colistin, and 0.25 to 2 μg/ml and 82.2% or 98.6% for tigecycline. However, when they were combined, the MIC_range values and susceptibility rates of MDR A. baumannii for nontraditional antibiotics were 0.25 to 4 μg/ml and 90.4% for rifampin, 0.0625 to 32 μg/ml and 26% for biapenem, 0.0625 to 0.5 μg/ml and 100% for colistin, and 0.125 to 1 μg/ml and 100% for tigecycline.

When rifampin, biapenem, colistin, and tigecycline were used alone, the percentages of the 73 MDR A. baumannii isolates whose MICs were lower than the maximum concentration of the drug (C_max) in serum were 83.1%, 30.1% to 76.7%, 100%, and 24.7% to 82.2%, respectively. However, when rifampin was used in combination (with either biapenem, colistin, or tigecycline), the percentages of isolates whose MICs were lower than the C_max in serum were 100%, 90.4% to 100%, 100%, and 91.8% to 100%, respectively (Table 1).

TABLE 1.

MICs of rifampin, biapenem, colistin, and tigecycline used alone and in combination against multidrug-resistant Acinetobacter baumannii^a

Antimicrobial	Data for the indicated antibiotic used alone					Data for the indicated antibiotic used in combinatione
	MIC50 (μg/ml)	MIC90 (μg/ml)	MICrangeb (μg/ml)	%Sc	% < Cmaxd	MIC50 (μg/ml)	MIC90 (μg/ml)	MIC range (μg/ml)	%S	% < Cmax
Rifampin	4	8	1–16	30	83.1	1	2	0.25–4	95.9/90.4/97.3	100
Biapenem	32	64	1–64	5.5	30.1–76.7	8	16	0.0625–32	26	90.4–100
Colistin	0.5	1	0.125–1	100	100	0.125	0.25	0.0625–0.5	100	100
Tigecycline	1	2	0.25–2	82.2/98.6	24.7–82.2	0.25	0.5	0.125–1	100	91.8–100

^an = 73.

^bMIC_range, from the minimum value of MICs to the maximum value of MICs.

^c%S, percent susceptible.

^d% < C_max, percentage of the tested drug's MIC at less than or equal to the peak concentration in serum.

^eRifampin was combined with biapenem, colistin, or tigecycline, and each of the other antimicrobials was combined with rifampin.

FICIs for rifampin combined with the other antimicrobials against MDR A. baumannii.

The results of the checkerboard synergy analysis of the MDR A. baumannii strains are shown in Table 2. Of the tested strains, 23 (31.51%), 36 (49.31%), and 14 (19.18%) showed synergistic (FICI ≤ 0.5), partially synergistic (0.5 < FICI < 1), and additive interactions (FICI = 1), and the percentages of isolates whose MICs were lower than the one-third C_max of biapenem in serum were 56.6% (FICI ≤ 0.5), 47.2% (0.5 < FICI < 1), and 28.5% (FICI = 1) between rifampin and biapenem, respectively. Twenty-five (34.25%), 32 (43.83%), and 16 isolates (21.92%) showed synergistic, partially synergistic, and additive interactions, and the percentages of isolates whose MICs were lower than the 1/16 C_max of colistin in serum were 92%, 68.7%, and 15.4% between rifampin and colistin, respectively. Twenty-three (31.51%), 35 (47.94%), and 15 isolates (20.55%) showed synergistic, partially synergistic, and additive interactions, and the percentages of isolates whose MICs were lower than the one-third C_max of tigecycline in serum were 91.3%, 45.7%, and 66.7% between rifampin and tigecycline, respectively. No indifferent (1 < FICI < 4) or antagonistic (FICI ≥ 4) interactions were observed (Table 2).

TABLE 2.

FICIs of rifampin combined with each of the three other antimicrobials against multidrug-resistant Acinetobacter baumannii^a

Standard of result judgment	Interaction	FICI of rifampin combined with:
		Biapenem		Colistin		Tigecycline
		No. (%)b	No. < 1/3Cmaxc	No. (%)	% < 1/16Cmaxd	No. (%)	% < 1/3Cmax
FICI ≤ 0.5	Synergistic	23 (31.51)	56.6	25 (34.25)	92	23 (31.51)	91.3
0.5 < FICI < 1	Partially synergistic	36 (49.31)	47.2	32 (43.83)	68.7	35 (47.94)	45.7
FICI = 1	Additive	14 (19.18)	28.5	16 (21.92)	15.4	15 (20.55)	66.7
1 < FICI < 4	Indifferent	0 (0.00)	NAe	0 (0.00)	NA	0 (0.00)	NA
FICI ≥ 4	Antagonistic	0 (0.00)	NA	0 (0.00)	NA	0 (0.00)	NA

^an = 73.

^bNo. (%), number of strains and percentage of the FICI values.

^c% < 1/3C_max, percentage of the tested drug's MIC at ≤1/3 of the peak concentration in serum.

^d% < 1/16C_max, percentage of the tested drug's MIC at ≤1/16 of the peak concentration in serum.

^eNA, not available.

CIRs of rifampin and/or other antimicrobials against MDR A. baumannii.

The CIRs of rifampin and/or other antimicrobials (biapenem, colistin, and tigecycline) against MDR A. baumannii are shown in Fig. 1, 2, and 3. The curves for the CIRs moved markedly to the left when two drugs (rifampin with biapenem, colistin, or tigecycline) were used together compared with each antimicrobial drug used alone.

FIG 1.

The cumulative inhibition ratios (CIRs) of rifampin and/or biapenem against multidrug-resistant Acinetobacter baumannii (n = 73). C, concentration; ◊, rifampin as monotherapy; △, biapenem as monotherapy; solid line/◼, rifampin (after rifampin and biapenem were combined); dotted line/◼, biapenem (after biapenem and rifampin were combined).

FIG 2.

The cumulative inhibition ratios (CIRs) of rifampin and/or colistin against multidrug-resistant Acinetobacter baumannii (n = 73). C, concentration; ◊, rifampin as monotherapy; △, colistin as monotherapy; dotted line/◼, rifampin (after rifampin and colistin were combined); solid line/◼, colistin (after colistin and rifampin were combined).

FIG 3.

The cumulative inhibition ratios (CIRs) of rifampin and/or tigecycline against multidrug-resistant Acinetobacter baumannii (n = 73). C, concentration; ◊, rifampin as monotherapy; △, tigecycline as monotherapy; solid line/◼, rifampin (after rifampin and tigecycline were combined); dotted line/◼, tigecycline (after tigecycline and rifampin were combined).

DISCUSSION

A. baumannii is a major nosocomial infectious agent and has become a problematic pathogen over the last decade (17). It displays extensive antibacterial resistance and recently acquired resistance to carbapenems, which were previously used as a salvage therapy (18). Confronted with the worldwide increase in MDR A. baumannii infections, the selection of effective drugs and combination therapy have become increasingly important in medical practice, and more combination treatments have recently been reported (19). Combining antimicrobials is an effective treatment approach for infection caused by MDR A. baumannii (20). A common reason for using combined antimicrobial therapy is synergistic interaction (20). Therefore, some banned and old drugs should be reassessed, and the investigation of novel combinations of existing agents is clearly warranted.

Some antimicrobials are unable to inhibit or kill bacteria, which may enhance the combined activity of other antimicrobials (21); some antimicrobials which are able to inhibit or kill bacteria by themselves may produce resistance rapidly or require a high dose and have serious toxicity. Some reports showed that resistance has been associated with monotherapy. For example, rifampin demonstrated in vitro and in vivo bactericidal activities against MDR A. baumannii in an experimental model of pneumonia, whereas rifampin-resistant mutants appeared shortly after the initiation of treatment (18, 22). Moreover, rifampin combined with imipenem or sulbactam was effective in the treatment of experimental pneumonia and meningitis caused by imipenem-resistant A. baumannii (23). Carbapenems are the antimicrobial agents of choice against serious infections of MDR A. baumannii, but resistance to these antimicrobial agents has been reported worldwide (23). Like other carbapenems, the structure of biapenem confers stability against most β-lactamases; moreover, it remains stable against extended-spectrum β-lactamases, so it is active against isolates that are already resistant to imipenem (24). However, no in vitro synergistic effects of biapenem combined with other antimicrobials have been reported. Colistin is considered the treatment of last resort, an established antimicrobial from the polymyxin family. It is effective against MDR A. baumannii isolates. However, the efficacy and toxicity of colistin and the potential concern that it may drive resistance when used as a monotherapy must be discussed (25). Colistin exerts a rapid bactericidal effect at high concentrations, whereas it affects the outer membrane and increases the permeability of Gram-negative bacteria at lower concentrations, facilitating the entry of excluded compounds (26). These compounds include hydrophobic drugs, such as macrolides and glycopeptides (27) (teicoplanin [28], telavancin [29], and daptomycin [30]), which do not affect Gram-negative bacteria but show potent synergistic activity when combined with colistin. Tigecycline is a new semisynthetic tetracycline for the treatment of MDR A. baumannii infections, including carbapenem-resistant isolates. However, resistance develops when isolates with reduced susceptibility are exposed to sub-MIC values (31), which together with its suboptimal pharmacokinetic and pharmacodynamic profiles strongly limits its clinical use (32). Therefore, the efficacy of rifampin against MDR A. baumannii when it is individually combined with biapenem, colistin, or tigecycline was evaluated.

In this study, 73 MDR A. baumannii isolates from two A-level general hospitals in Beijing were used. Rifampin combined with biapenem, colistin, or tigecycline was synergistic with 31.51%, 34.25%, and 31.51%, partially synergistic with 49.31%, 43.83%, and 47.94%, additive with 19.18%, 21.92%, and 20.55%, and indifferent and antagonistic with 0% of the strains, respectively. Significant in vitro synergistic and partially synergistic interactions were observed. Most of the isolates were susceptible to rifampin and biapenem, but all the included isolates were susceptible to colistin, and 82.2% or 98.6% of the isolates were susceptible to tigecycline. However, in combinations, rifampin MIC values for 90.4% of the strains were ≤2 μg/ml, biapenem MIC values for 90.4% or 100% of the strains were ≤4 μg/ml, colistin MIC values for 100% of the strains were ≤1 μg/ml, and tigecycline MIC values for 100% of the strains were ≤0.5 μg/ml. Moreover, the mean serum C_max of rifampin is 18.76 ± 6.61 μg/ml when it is used for the treatment of Mycobacterium avium complex lung disease by mean daily dose (9.62 ± 1.63 mg/kg of body weight) (33), the mean serum C_max of biapenem is 24.7 ± 4.97 μg/ml by intravenous infusion (300 mg) administered three times daily (34), the serum C_max of colistin is 2 μg/ml (occurring 3 h after an initial dose of 2 million IU of colistimethate sodium [35]), and the mean serum C_max of tigecycline is 0.72 ± 0.24 μg/ml in a normal dose due to the large volume of distribution (36). Despite the outcome of tigecycline treatment, in vitro activity is exciting. Among the 73 MDR A. baumannii isolates, 100% of the isolates had MICs lower than the C_max values of rifampin, biapenem, colistin, and tigecycline in serum when used in combination. Moreover, the MIC₅₀, MIC₉₀, and MIC_range against the isolates of rifampin combined with biapenem, colistin, and tigecycline were clearly lower than those for each antimicrobial alone. Before combination, the MIC₅₀, MIC₉₀, and MIC_range values of rifampin were 4, 8, and 1 to 16 μg/ml; for biapenem, they were 32, 64, and 1 to 64 μg/ml, respectively. When they were combined, the MIC₅₀, MIC₉₀, and MIC_range values against the isolates were decreased by more than 4 times for each agent. Similar results have been achieved when rifampin was combined with colistin and tigecycline. From in vitro results, the combinations of rifampin with biapenem, colistin, and tigecycline were active at lower concentrations than those for the monotherapy. These combinations may also prevent the emergence of resistance to current MDR A. baumannii drugs. Hence, rifampin combined with biapenem, colistin, or tigecycline should be used when treating MDR A. baumannii infections to prevent the emergence of resistance to either of these antimicrobials. The percentage of the tested drug MICs that were less than or equal to the peak concentration in serum (%MIC < C_max) and the CIR curve can still show the notable difference in combination therapy, supporting combination therapy as a future alternative which may provide optimized pharmacokinetics for each drug.

In conclusion, our in vitro results suggest that the combinations of rifampin with biapenem, colistin, or tigecycline have synergistic and partially synergistic effects against MDR A. baumannii isolates. The presence of rifampin, biapenem, colistin, and tigecycline is crucial to prevent further emergence of these nonsusceptible strains. For this reason, in vitro activity of combination therapy of rifampin with biapenem, colistin, and tigecycline against MDR A. baumannii has been investigated; such combinations may be useful in the treatment of infections caused by MDR A. baumannii. These results may be verified with a more appropriate methodology for further studies, especially time-kill and in vivo synergy studies. Animal models can be considered to establish an in vivo study; hence, the synergistic activity can be tested in vivo with these combinations.