In vitro efficacy of various antibiotic combinations against Pseudomonas aeruginosa isolates

Abstract

Objective

Pseudomonas aeruginosa is one of the leading causes of nosocomial infection. The present study tested the in vitro efficacy of ceftazidime or imipenem combined with amikacin, levofloxacin and colistin in P.aeruginosa isolates.

Methods

P.aeruginosa strains, isolated from clinical samples, were assessed for antibiotic susceptibility using the disc diffusion method. Antibiotic combination tests were performed using minimum inhibitory concentration (MIC) test strips and the sum of the Fractional Inhibitory Concentration (ΣFIC) index was used to assess synergy.

Results

Out of 60 isolated P.aeruginosa strains, 100% were susceptible to colistin and 26.7% (16 strains) were multidrug resistant. MIC50 and MIC90 values were 2 and 32 µg/ml for imipenem; 1.5 and 24 µg/ml for ceftazidime; 3 and 8 µg/ml for amikacin; 0.38 and 32 µg/ml for levofloxacin; 1 and 1.5 µg/ml for colistin, respectively. Antagonism was not found in any of the antibiotic combinations tested. The amikacin–ceftazidime combination was found to have a synergistic effect in 15% of the strains, but no synergistic effect was detected for other combinations.

Conclusions

In Pseudomonas infection, alternative treatment options using different antibiotic combinations should be tested in vitro and findings should be confirmed by clinical studies.

Introduction

Pseudomonas aeruginosa is a remarkable bacterium due to its ability to grow in minimally favourable conditions, its widespread presence in nature, its natural and acquired resistance mechanisms, and its multiple virulence factors (with various enzymes and toxins). ¹ P. aeruginosa may readily colonize the hospital setting and is one of the most prevalent pathogens responsible for nosocomial infections. ² The various resistance mechanisms of P. aeruginosa include enzymatic hydrolysis (β-lactamases, particularly metallo-β-lactamase), efflux pump and low intrinsic outer membrane permeability. ³ As a result, infection with multidrug-resistant Pseudomonas causes serious problems in the treatment of nosocomial infection, and resistant P. aeruginosa infections increase mortality, morbidity and treatment costs. ⁴ Alternative drugs for the treatment of those Pseudomonas infections that are resistant to multiple antibiotics (resistant to more than three groups) have yet to be investigated. ³

Concomitant use of antibiotics (combination therapy) is recommended for severe infections when P. aeruginosa is the suspected pathogen, in order to prevent the development of resistance during treatment and to achieve a wide spectrum of activity. In addition to preventing the development of resistance, combined use of antibiotics (such as β-lactams, quinolones and aminoglycosides) may have synergistic effects and may reduce the occurrence of side-effects, since each drug is used at a lower dose than would be used for monotherapy. ⁵

Susceptibility tests to investigate the synergistic antibacterial effectiveness of two different drugs for the same pathogen can be difficult to perform in vitro, and drugs are usually chosen according to the relevant guidelines. It is therefore important to ensure guidelines are constantly updated with information on the changing resistance rates.

In the present study, various antibiotic combinations were tested in vitro for their effectiveness against P. aeruginosa, with the aim of updating information on resistance rates for several widely used antibiotic combinations, and to provide data on the effectiveness of new combinations. Guideline-recommended combinations, ⁶ in addition to novel combinations including colistin (a recently introduced drug that is underutilized in combination tests), were used in the present study.

Materials and methods

Identification of bacteria

This study included P. aeruginosa strains isolated from hospitalized patients treated at The Faculty of Medicine, Gaziantep University, Sahinbey Research and Training Hospital, Gaziantep, Turkey, between December 2012 and June 2013. No investigation was undertaken regarding the origins of each pathogen and whether or not the infections were nosocomial. This study was approved by the Ethics Committee of the Faculty of Medicine, Gaziantep University, Turkey (No. 04.12.2012/448).

Clinical isolates were characterized using conventional methods and identified using a BD Phoenix™ Automated Microbiology System (BD, San Jose, CA, USA). Isolates were stored at –80℃ and subcultured in brain–heart infusion agar (Oxoid, Basingstoke, UK) prior to testing.

Antibiotic susceptibility testing

Antibiotic susceptibility tests were performed using the disc diffusion method according to Clinical and Laboratory Standards Institute (CLSI) guidelines. ⁷ Imipenem, meropenem, aztreonam, ceftazidime, cefepime, gentamicin, amikacin, tobramycin, ciprofloxacin, levofloxacin, piperacillin/tazobactam, cefoperazone/sulbactam, and ticarcillin/clavulonic acid antibiotic discs (Oxoid) were used for testing.

Susceptibility to colistin was determined by Liofilchem® minimum inhibitory concentration (MIC) test strips (Liofilchem, Abruzzo, TE, Italy), since colistin is not commercially available in disc form. P. aeruginosa ATCC®27853™ (ATCC, Manassas, VA, USA) was used as a susceptible control strain.

Antibiotic combination testing (synergy test)

Antibiotic combinations were tested using the Liofilchem® MIC test strips, to determine the combined antimicrobial effects on the isolated P. aeruginosa strains. ⁸ Liofilchem® MIC test strips had the following standard antibiotic concentrations (µg/ml): colistin, 0.016–256; amikacin, 0.016–256; ceftazidime, 0.016–256; imipenem, 0.002–32; and levofloxacin, 0.002–32.

The following antibiotic combinations were tested: amikacin + ceftazidime; amikacin + imipenem; levofloxacin + ceftazidime; levofloxacin + imipenem; colistin + ceftazidime; colistin + imipenem. MIC required to inhibit 50% and 90% of the strains (MIC50 and MIC90, respectively) were calculated using Microsoft® Excel® 2007 software (Microsoft Corp., Redmond, WA, USA).

Bacterial suspensions homogenized in sterile saline were prepared from overnight fresh cultures to a McFarland standard of 0.5 and inoculated into 150 mm Mueller-Hinton agar plates (four plates per bacterial isolate). The agar plates were allowed to stand for 15–20 min at room temperature to allow any excess surface moisture to be absorbed before placement of MIC test strips according to the following protocol: For each bacterial isolate, three amikacin strips were placed onto the first plate; three levofloxacin strips were placed onto the second plate; three colistin strips were placed onto the third plate; and one strip each of imipenem and ceftazidime was placed onto the fourth plate (to determine MIC for imipenem and ceftazidime). Strips were placed at 30-mm intervals.

Plate numbers 1, 2 and 3 were used for combination testing. For each plate, the first two strips were replaced with the second antibiotic strip according to the combination to be tested. The third strip was left in place to determine the MIC value of the initial antibiotic alone.

The start- and endpoints of strips to be replaced were marked on the back of each plate. Following incubation for 1 h at room temperature, amikacin, levofloxacin and colistin MIC test strips were removed and replaced with imipenem and ceftazidime strips, ensuring that the concentration lines overlapped. The plates were then incubated at 37℃ for 18–20 h, after which the antimicrobial concentration at the point where the diameter of the inhibition zone intersected with the edge of the MIC test strip was recorded as the MIC value. MIC values were interpreted as recommended by CLSI guidelines. ⁷

Estimation of FIC index

The sum of the fractional inhibitory concentration index (ΣFIC) was calculated to determine the effect of antibiotic combinations. For antibiotics used in combination, the first antibiotic was designated ‘A’ and the second ‘B’. The ΣFIC value was calculated using the following formula:

$\Sigma \text{FICindex}={\text{FIC}}_{\text{A}}+{\text{FIC}}_{\text{B}}$

${\text{FIC}}_{\text{A}}=\text{combined}{\text{MIC}}_{(\text{A}+\text{B})}/{\text{MIC}}_{\text{A}}$

${\text{FIC}}_{\text{B}}=\text{combined}{\text{MIC}}_{(\text{A}+\text{B})}/{\text{MIC}}_{\text{B}}$

Results of calculations were interpreted as previously described. ⁹ A calculated ΣFIC value of ≤0.5 represented a synergistic effect (i.e. total effect greater than the sum of the individual antibiotic effects), a value between >0.5 and <2 represented an indifferent (additive) effect (i.e. no additional contribution from including the second antibiotic, compared with use of the first antibiotic alone), and a value of ≥2 represented an antagonistic effect (i.e. total effect less than the sum of the individual effects).

Results

Sixty P. aeruginosa strains isolated from various clinical samples were included in the present study (Table 1) and all (100%) were found to be susceptible to colistin (Table 2). The second most effective antibiotic was amikacin, to which 56 isolates (93.3%) were susceptible. The least effective antibiotic was aztreonam (26 susceptible isolates [43.3%]).

Table 1.

Distribution of 60 clinically isolated Pseudomonas aeruginosa strains according to hospital department and sample type of origin.

Hospital department	Sample type
	Wound swab	Tracheal aspirate	Urine	Sputum	BAL	Blood culture	Catheter	Total
Internal medicine	7	–	2	2	–	–	1	12 (20.0)
Internal medicine ICU	–	7	–	–	–	1	–	8 (13.3)
Chest diseases	–	–	–	6	1	–	–	7 (11.7)
Reanimation	–	4	–	–	–	2	–	6 (10.0)
Orthopaedics	6	–	–	–	–	–	–	6 (10.0)
Paediatrics	1	–	3	–	–	–	–	4 (6.7)
Chest surgery	–	–	–	–	4	–	–	4 (6.7)
Urology	–	–	3	–	–	–	–	3 (5.0)
Paediatric ICU	1	1	–	–	–	–	–	2 (3.3)
Other a	4	2	1	–	–	1	–	8 (13.3)
Total	19 (31.7)	14 (23.3)	9 (15.0)	8 (13.3)	5 (8.3)	4 (6.7)	1 (1.7)	60 (100)

Data presented as n (%) of Pseudomonas aeruginosa isolates.

BAL, bronchoalveolar lavage; ICU, intensive care unit.

a

Neurosurgery, Physical therapy and rehabilitation, Otolaryngology, Plastic surgery, Coronary ICU, General ICU, Chest Surgery ICU.

Table 2.

Antibiotic susceptibility of 60 clinically isolated Pseudomonas aeruginosa strains tested using the disc diffusion method.

Drug	Susceptible isolates
Amikacin	56 (93.3)
Aztreonam	26 (43.3)
Cefepime	47 (78.3)
Cefoperazone–sulbactam	32 (53.3)
Ciprofloxacin	46 (76.7)
Colistin a	60 (100.0)
Levofloxacin	44 (73.3)
Meropenem	39 (65.0)
Imipenem	39 (65.0)
Piperacillin–tazobactam	48 (80.0)
Ceftazidime	48 (80.0)
Gentamicin	46 (76.7)
Ticarcillin–clavulanate	28 (46.7)
Tobramycin	49 (81.7)

Data presented as n (%) of Pseudomonas aeruginosa isolates.

a

Tested using a minimum inhibitory concentration test-strip method.

A total of 16 P. aeruginosa isolates (26.7%) were susceptible to all of the antibiotics tested. Another 16 (26.7%) strains were found to be resistant (multidrug resistant) to at least three classes of antibiotics including cephalosporins (only ceftazidime or cefepime), aminoglycosides, fluoroquinolones, carbapenems and piperacillin.

Synergy tests

The MIC ranges for imipenem, ceftazidime, amikacin, levofloxacin and colistin tested against 60 isolated P. aeruginosa strains, and the MIC50 and MIC90 values, are shown (Table 3).

Table 3.

Minimum inhibitory concentration (MIC) values for colistin, amikacin, levofloxacin, imipenem and ceftazidime, tested against 60 clinically isolated Pseudomonas aeruginosa strains.

Antibiotic	MIC range µg/ml	MIC50 µg/ml	MIC90 µg/ml
Imipenem	0.5–32	2	32
Ceftazidime	0.25–256	1.5	24
Amikacin	0.5–256	3	8
Levofloxacin	0.125–32	0.38	32
Colistin	0.75–2	1	1.5

The effects of antibiotic combinations tested against the 60 clinically isolated P. aeruginosa strains were assessed using the ΣFIC index criteria (Table 4). Three hundred and sixty FIC values were obtained for six antibiotic combinations tested on 60 P. aeruginosa isolates; of these, 351 (97.5%) were indifferent and nine (2.5%) exhibited synergistic effects. Antagonism was not found between any of the antibiotic combinations tested. Amikacin–ceftazidime was the only combination to show synergistic effects (seen in nine isolates, 15%). Among the 16 (26.7%) multidrug-resistant isolates, the amikacin–ceftazidime combination showed a synergistic effect in one isolate (6.3%; data not shown).

Table 4.

Effects of antibiotic combinations tested against 60 clinically isolated Pseudomonas aeruginosa strains assessed using minimum inhibitory concentration test strips.

	Test results
Antibiotic combination	Synergistic effect	Indifferent effect	Antagonistic effect
Colistin–imipenem	0 (0)	60 (100)	0 (0)
Colistin–ceftazidime	0 (0)	60 (100)	0 (0)
Amikacin–imipenem	0 (0)	60 (100)	0 (0)
Amikacin–ceftazidime	9 (15)	51 (85)	0 (0)
Levofloxacin–imipenem	0 (0)	60 (100)	0 (0)
Levofloxacin–ceftazidime	0 (0)	60 (100)	0 (0)

.

Data presented as n (%) of Pseudomonas aeruginosa isolates

Discussion

Pseudomonas aeruginosa is one of the most prevalent pathogens found in the hospital environment. ¹⁰ P. aeruginosa is an opportunistic pathogen which has the ability to survive in harsh conditions; its presence may lead to severe, difficult-to-treat infections in immunocompromised patient populations including patients with cystic fibrosis, burn-related infections or HIV-positivity. ¹¹

The increased observance of multidrug resistance (mainly to β-lactam antibiotics) in Pseudomonas strains isolated from nosocomial infections is making it increasingly difficult to treat infections caused by this pathogen. ² Resistance to antimicrobials in Pseudomonas strains develops via a number of mechanisms, including the production of specific enzymes (β-lactamases, enzymes that modify aminoglycosides, for example), changes in cell-membrane permeability and active efflux systems. ¹²

The rate of antibiotic resistance in the hospital setting depends mainly on the antibiotic policy, but also on the physical structure of that hospital, patient characteristics, and frequency of invasive procedures performed. ¹³ Knowledge of the resistance phenotype and local resistance surveillance is important for empirical treatment. ⁶

Numerous studies relating to the antibiotic susceptibility of P. aeruginosa have been published.^14–18 In the present study, none of the tested isolates were resistant to colistin, which was found to be the most effective antibiotic. Amikacin was found to be the second most-effective antibiotic, and of the 60 isolates tested, 16 (26.7%) were multidrug resistant.

Resistance to an antibiotic used for treatment more commonly develops in P. aeruginosa infections compared with other bacteria, and antibiotic monotherapy is frequently associated with treatment difficulties compared with combination regimens. ² Thus, combined antibiotic therapy is usually administered for treating P. aeruginosa infections, in order to achieve powerful antibacterial efficacy, to slow development of resistance and to reduce the possibility of side-effects by giving lower doses of individual drugs.^5,19

The most commonly used antibiotic combination in P. aeruginosa infections is the β-lactam/aminoglycoside combination, which usually shows a synergistic effect.^20,21 Several different drug combinations have been previously tested using various investigation techniques,^19,21–23 each with their own advantages and disadvantages. In the present study, the MIC test strip method was chosen for ease of performance and evaluation, and because of ready access to the test strips relating to the antibiotics undergoing investigation.

In the present study, amikacin–ceftazidime and amikacin–imipenem combinations were tested to explore the synergistic effects of an aminoglycoside/β-lactam combination, but such an effect was only observed for amikacin–ceftazidime (in 15% of isolates). Synergistic interactions involving these combinations have been reported in many studies.^22,24,25

Use of β-lactam and fluoroquinolone in combination (ciprofloxacin/levofloxacin) has increased considerably, and synergistic effects have been reported. ²⁰ A study of P. aeruginosa found a synergistic effect for imipenem–ciprofloxacin in 3.1% of strains, no synergistic effect for imipenem–levofloxacin, a synergistic effect for meropenem–ciprofloxacin in 6.2% of strains, and no synergistic effect for meropenem–levofloxacin. ²⁶ In the present study, levofloxacin was chosen to represent quinolones and the efficacies of levofloxacin–ceftazidime and levofloxacin–imipenem combinations were investigated. No synergistic effect was found for levofloxacin–ceftazidime in any of the strains tested. Similarly, there was no synergistic effect for the levofloxacin–imipenem combination. These results suggest that an antibiotic combination involving levofloxacin may not be the most suitable first-line treatment option in the hospital setting as used in the present study.

Colistin is regaining popularity as a treatment option for multidrug-resistant infections; in vitro and limited clinical studies, as well as its use as monotherapy, suggest that colistin may also be used in combination with other drugs (carbapenems, other β-lactams, aminoglycosides and quinolones). ²⁷ Several studies have tested colistin in combination with imipenem, meropenem, ceftazidime and rifampicin and reported either synergistic effects or ineffectiveness.^28,29

In the present study, the colistin–imipenem and colistin–ceftazidime combinations both showed indifferent effects in 100% of strains tested, however no antagonistic effect was observed. This might have resulted from the fact that all of the studied isolates were already susceptible to colistin, which suggests that this antibiotic should not be the first choice for monotherapy.

Out of 16 multidrug-resistant isolates tested in the present study, the amikacin–ceftazidime combination showed a synergistic effect in one isolate (6.3%). Thus, aminoglycoside/cephalosporin antibiotic combinations may represent potential first-line options for treating multidrug resistant isolates, when limited by treatment options. Susceptibility testing should be properly analysed during antibiotic selection and individual hospitals should select the antibiotic regimen that is most suitable for the established resistance profile within the hospital.

The present study was limited by budget restrictions. Disc-diffusion and MIC-strip testing is costly to undertake; consequently only 60 consecutive clinically isolated Pseudomonas strains were included, and combinations of colistin, amikacin and levofloxacin with ceftazidime and imipenem were investigated, governed by the number of MIC test strips available. Ideally, the present in vitro results should also be supported by in vivo studies.

In conclusion, we suggest that use of an appropriate antibiotic for treatment of P. aeruginosa infection is more important than the discovery of a novel antibiotic, in relation to managing bacterial resistance mechanisms. Susceptibility testing should be properly analysed during antibiotic selection and individual hospitals should select the antibiotic regimen that is most suitable for the established resistance profile within the hospital.