In Vitro Bactericidal Activity of Trimethoprim-Sulfamethoxazole Alone and in Combination with Colistin against Carbapenem-Resistant Acinetobacter baumannii Clinical Isolates

Abstract

Trimethoprim-sulfamethoxazole alone and combined with colistin was tested in vitro against six carbapenem-resistant Acinetobacter baumannii (CRAB) clinical strains. After 24 h, at achievable serum concentrations, trimethoprim-sulfamethoxazole effectively killed all strains, while colistin killed only one strain. Trimethoprim-sulfamethoxazole plus colistin rapidly killed all strains after 6 h and for up to 24 h. Trimethoprim-sulfamethoxazole, one of the few remaining antimicrobials that still has a degree of activity, particularly combined with colistin, might represent an effective therapy for severe CRAB infections.

TEXT

Acinetobacter baumannii has been an important nosocomial pathogen for the past 30 years, being frequently implicated in ventilator-associated pneumonia and bloodstream, urinary tract, and soft tissue infections (1) and having a propensity to develop particular antibiotic resistance. According to the ECDC's annual epidemiological surveillance report for 2014, carbapenem-resistant A. baumannii (CRAB) is endemic in many European countries; half of the reported isolates from all over Europe were resistant to carbapenems, aminoglycosides, and fluoroquinolones, while in Greece, the respective percentages were 93%, 89%, and 95% (2).

Trimethoprim-sulfamethoxazole is a combination of two antimicrobial agents that act synergistically against a wide variety of bacteria. Maximum synergistic inhibition occurs when the trimethoprim/sulfamethoxazole concentration ratio is 1:20. The drug combination is prepared at a fixed ratio of 1:5; however, the achievable serum concentration after both oral and intravenous administration is 1:20, because of the wider volume of distribution of trimethoprim than of sulfamethoxazole. The combination may be effective in a variety of infections, such as respiratory, urinary, gastrointestinal, and skin and wound infections and septicemias. Τtrimethoprim and sulfamethoxazole exert their synergistic effect by inhibiting successive steps in the folate synthesis pathway (3). Despite the scarcity of novel antibiotics active against carbapenem-resistant Gram-negative bacteria and the existing need to rely on available old antibiotics, the in vitro killing activity of trimethoprim-sulfamethoxazole against CRAB has not been studied.

During the last few years, susceptibility data for A. baumannii from our large tertiary hospital have shown high rates of resistance to carbapenems (>90% for imipenem and meropenem), while resistance to trimethoprim-sulfamethoxazole has been lower (approximately 70%) (unpublished data). The aim of our study was to test the in vitro antibacterial activity of trimethoprim-sulfamethoxazole, alone and in combination with colistin, against CRAB blood isolates. Trimethoprim-sulfamethoxazole is an old drug, and a large body of pharmacokinetic/pharmacodynamic data is available, including phase IV data on safety profiles. In that respect, the results of this study might be readily applicable by clinicians treating CRAB infections, possibly broadening their therapeutic arsenal.

Bacterial strains and susceptibility testing.

During the period from February 2011 to September 2013, 48 single-patient trimethoprim-sulfamethoxazole-susceptible A. baumannii clinical isolates were recovered from blood cultures of patients hospitalized in Evaggelismos, a 1,000-bed tertiary-care hospital with 40 intensive care unit (ICU) beds. Identification and initial susceptibility testing were performed using the Vitek2 system (bioMérieux, Marcy l'Etoile, France). Of these isolates, six trimethoprim-sulfamethoxazole- and colistin-susceptible CRAB strains (strains 1, 14, 16, 30, 38, and 45) (Table 1) were selected for further study. The MICs of these strains for trimethoprim, sulfamethoxazole, trimethoprim-sulfamethoxazole (at a ratio of 1:19), and colistin were determined by broth macrodilution (4), which is the commonly applied standard method for isolates to be tested by time-kill assays. The EUCAST clinical breakpoints were used for colistin (susceptible, ≤2 μg/ml; resistant, >2 μg/ml) and trimethoprim-sulfamethoxazole (susceptible, ≤2 μg/ml; resistant, >4 μg/ml [breakpoints are expressed as trimethoprim concentration]) (5). As neither EUCAST nor CLSI have breakpoints for trimethoprim or sulfamethoxazole versus A. baumannii, we applied the CLSI resistance breakpoints for Enterobacteriaceae (for trimethoprim [susceptible, ≤8 μg/ml; resistant, ≥16 μg/ml] and sulfamethoxazole [susceptible, ≤256 μg/ml; resistant, ≥512 μg/ml]) (5, 6). Escherichia coli ATCC 25922 and A. baumannii ATCC 19606 were used as quality controls in all susceptibility assays.

TABLE 1.

Characteristics of the six strains tested and results of bactericidal and synergy assays

Strain	Carbapenemase gene	PFGE type	MIC (μg/ml)				Mean change (log10 CFU/ml) from initial bacterial concn after incubation for 24 h
			Trimethoprim	Sulfamethoxazole	Trimethoprim-sulfamethoxazole	Colistin	Trimethoprim-sulfamethoxazole			Colistin		Trimethoprim-sulfamethoxazole + colistin
							1× MIC	2 μg/ml	4 μg/ml	1× MIC	2 μg/ml	1× MIC	2 μg/ml
1	blaOXA-58	II	32	256	0.25	0.5	0.01	−3.19a	−4.69	0.31	0.31	0.01	−4.69
14	blaOXA-23	I	256	>1,024	1	0.5	0.01	0.01	−3.52	0.21	0.31	−0.29	−3.92
16	blaOXA-23	I	256	512	1	0.5	−0.19b	−1.99	−4.69	0.31	−2.69	0.01	−4.69
30	blaOXA-23	VII	16	512	0.25	0.5	0.31	−3.09	−4.69	0.31	0.48	−0.29	−4.69
38	blaOXA-23	I	64	256	0.5	1	0.21	−3.09	−4.69	0.31	−4.69	0.31	−4.69
45	blaOXA-23	I	256	>1,024	1	1	−0.49	−3.39	−3.39	0.21	0.31	−3.39c	−4.69

^aA ≥3-log₁₀ reduction in CFU/ml implies a bactericidal effect.

^bA <3-log₁₀ reduction in CFU/ml implies a bacteriostatic effect.

^cA ≥2-log₁₀ reduction in CFU/ml of a drug combination at 24 h compared with the result for the most active drug implies synergism.

Molecular testing.

The strains were confirmed to be A. baumannii by PCR typing of bla_OXA-51-like (7). Epidemiological typing was carried out by pulsed-field gel electrophoresis (PFGE). Molecular assays were performed for common OXA-type carbapenemase (bla_OXA-23-type, bla_OXA-40-type, and bla_OXA-58-type) and metallo-β-lactamase (bla_VIM and bla_IMP) genes (8).

Bactericidal and synergy assays.

The bactericidal activities of trimethoprim-sulfamethoxazole and colistin were determined by time-kill assays. These assays were performed in triplicate, with inoculation of 5 × 10⁵ CFU/ml of each strain into 3 ml of fresh cation-adjusted Mueller-Hinton broth. The applied concentrations for trimethoprim-sulfamethoxazole were 1× MIC, 2 μg/ml, and 4 μg/ml, and those for colistin were 1× MIC and 2 μg/ml. These concentrations were selected because 4 and 2 μg/ml are the highest achievable levels in serum for trimethoprim-sulfamethoxazole and colistin, respectively (9–12). Synergy testing using the combination trimethoprim-sulfamethoxazole plus colistin was performed with the above-mentioned six strains, using antibiotic concentrations of 1× MIC for each compound and also 2 μg/ml for both drugs. Aliquots were removed after incubation for 0, 6, 12, 18, and 24 h, serially diluted, and plated on Mueller-Hinton agar plates to enumerate viable colonies; the lower limit of detection was 10 CFU/ml.

Bactericidal activity was defined as a ≥3-log₁₀ reduction in the total number of CFU/ml compared with the level in the initial inoculum, and bacteriostatic activity was defined as maintenance of or a <3-log₁₀ reduction in the total number of CFU/ml. Synergistic activity was defined as a ≥2-log₁₀ decrease in CFU/ml caused by the combination compared with the result for the most active antibiotic, and indifference was defined as a ±1-log₁₀ to <2-log₁₀ reduction compared with the result for the most efficient compound alone (4). Time-kill curves were prepared by plotting mean colony counts versus time.

Susceptibility results.

Of the 48 trimethoprim-sulfamethoxazole-susceptible A. baumannii isolates, 43 (89.6%) were CRAB. The MICs for the six CRAB strains tested by time-kill assays are shown in Table 1; of note, the trimethoprim-sulfamethoxazole MICs were relatively low, ranging from 0.25 to 1 μg/ml. The colistin MICs were 0.5 to 1 μg/ml. Sulfamethoxazole alone was active against two strains (MICs ranged from 256 to >1,024 μg/ml), while all strains were resistant to trimethoprim (MICs ranged from 16 to 256 μg/ml), as expected due to the intrinsic resistance of A. baumannii to trimethoprim (4).

Typing results.

The isolates were classified into eight PFGE types (types I to VIII). Type I predominated, with 40 isolates; each of the other seven types comprised 1 or 2 isolates. The CRAB isolates belonged to type I (all 40 isolates were CRAB harboring OXA-23 carbapenemase gene), type II (2 isolates, OXA-58), and type VII (1 isolate, OXA-23). Isolates of types III to VI and VIII were carbapenem susceptible, with no carbapenemase genes.

Bactericidal activity of trimethoprim-sulfamethoxazole and colistin.

The six CRAB strains that were subjected to bactericidal assays were selected to represent PFGE types I, II, and VII (strains 14, 1, and 30, respectively), and strains 16, 38, and 45 were selected randomly from type I, the predominant type. The killing curves are shown in Fig. 1. After 24 h of incubation, all six strains were effectively killed at 4 μg/ml trimethoprim-sulfamethoxazole, and four were killed at 2 μg/ml. At 1× MIC of trimethoprim-sulfamethoxazole, bactericidal activity was not observed, while the drug was bacteriostatic against two isolates. Colistin showed no activity at 1× MIC, while at 2 μg/ml the drug was bactericidal for all strains at the first 6 h but then regrowth was observed in five isolates.

FIG 1.

Bactericidal assays using trimethoprim-sulfamethoxazole and colistin against the six CRAB study strains (strains 1, 14, 16, 30, 38, and 45). y axis, log₁₀ CFU/ml; x axis, time in hours; 1a, trimethoprim-sulfamethoxazole alone at 1× MIC; 1b, trimethoprim-sulfamethoxazole at 2 μg/ml; 1c, trimethoprim-sulfamethoxazole at 4 μg/ml; 1d, colistin alone at 1× MIC; 1e, colistin at 2 μg/ml; 1f, synergy assays using trimethoprim-sulfamethoxazole plus colistin at 1× MIC for both drugs; 1g, synergy assays using trimethoprim-sulfamethoxazole plus colistin at 2 μg/ml for both drugs.

Synergy assays of trimethoprim-sulfamethoxazole combined with colistin.

Trimethoprim-sulfamethoxazole plus colistin exhibited strong synergistic and bactericidal activity against all six CRAB strains tested when concentrations of 2 μg/ml were used for both drugs. At 1× MIC for both drugs, only one strain was effectively killed, while against the remaining five strains the combination was mainly indifferent (Fig. 1).

Antibiotics active against A. baumannii in Southern Europe are commonly limited to colistin and tigecycline (2). However, in recent years, resistance to these antimicrobials has also appeared (13, 14). As new antimicrobials are limited, clinicians had to rely on old ones, including trimethoprim-sulfamethoxazole, which in a recent review exhibited highly variable antibacterial activity, ranging from 0 to 94%, against multidrug-resistant A. baumannii (15). Interestingly, in Greece in 2014, trimethoprim-sulfamethoxazole was the most active antibiotic among those reported in the national resistance surveillance data (data for colistin, minocycline, and tigecycline were not included), retaining activity against 30% of A. baumannii blood isolates from Greek ICUs (http://www.mednet.gr/whonet/). It is thus evident that trimethoprim-sulfamethoxazole might be an option for the treatment of a considerable proportion of CRAB infections, at least in combination with other drugs.

The bactericidal activity of trimethoprim-sulfamethoxazole observed when the combination was used as a single drug was effective against all tested strains, using concentrations that can be achieved in serum. However, the treatment of severe CRAB infections regularly requires antibiotic combinations, most of those including colistin (14, 16, 17). The combination of colistin plus trimethoprim-sulfamethoxazole is studied here by killing assays for the first time with trimethoprim-sulfamethoxazole-susceptible CRAB, against which this combination could be used in clinical practice. Only one previous study existed, but that study tested only trimethoprim-sulfamethoxazole-resistant A. baumannii; the study implied a degree of synergy between colistin and trimethoprim-sulfamethoxazole (18). Interestingly, in our study, the combination of these two drugs was rapidly bactericidal against all strains tested and the killing effect was retained for 24 h, with no regrowth occurring. Further, synergy was observed in 5/6 (83.3%) isolates.

In conclusion, we believe that the observed in vitro bactericidal and synergistic combination of trimethoprim-sulfamethoxazole with colistin might be promising for infections caused by CRAB isolates susceptible to these antibiotics. However, clinical studies are warranted to better assess the performance of trimethoprim-sulfamethoxazole combinations against CRAB infections.

Funding Statement

This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.