Antibiotic susceptibility of 65 isolates of Burkholderia pseudomallei and Burkholderia mallei to 35 antimicrobial agents

Abstract

Objectives: Fifty isolates of Burkholderia pseudomallei and 15 isolates of Burkholderia mallei were tested for their susceptibilities to 35 antimicrobial agents, including agents not previously tested against these bacteria.

Methods: MICs were determined by agar dilution in Mueller–Hinton medium.

Results: Among the antibiotics tested, lower MICs were obtained with imipenem, ceftazidime, piperacillin, piperacillin/tazobactam, doxycycline and minocycline. Fluoroquinolones and aminoglycosides had poor activities. A single clinical isolate of B. pseudomallei was resistant to ceftazidime, co-amoxiclav and doxycycline but remained susceptible to imipenem.

Conclusions: Although B. mallei MICs are often lower, the overall results underline the importance of resistance in both species. The susceptibilities measured are consistent with the current recommendations for the treatment of B. pseudomallei and B. mallei infections.

melioidosis, glanders, MICs, resistance, biowarfare

Introduction

Burkholderia pseudomallei is responsible for melioidosis, which is considered a major community-acquired infection in endemic areas, leading the disease to be recognized as a medical problem in the armies involved in conflicts in Asia.¹ Because relapses of melioidosis are common, clinical management requires long courses of chemotherapy.¹Burkholderia mallei is the causative agent of glanders and primarily infects horses, mules and donkeys. It is also able to infect humans; intentional releases of this bacterium have been documented.²

B. pseudomallei and B. mallei are considered as potential biological warfare or bioterrorism agents and have been included in the B list of the CDC.³B. mallei and B. pseudomallei are intrinsically resistant to a wide range of antimicrobial agents including β-lactam antibiotics, aminoglycosides and macrolides.^4,5 However, few antibiotic susceptibility studies of B. mallei have been performed.^6,7 As there is a need for effective treatments and post-exposure prophylaxis, the objective of this study was to assess the in vitro susceptibilities of a large panel of strains of B. mallei and B. pseudomallei to a wide variety of antibiotics.

Materials and methods

Fifty isolates of B. pseudomallei and 15 isolates of B. mallei belonging to the collection of the Centre de Recherches du Service de Santé des Armées Emile Pardé (La Tronche, France) were included in this protocol. Strains were stored at −80°C in 20% glycerol. Strain origin is reported in Table 1. Species identification was confirmed by routine phenotypic characterization including Gram staining, motility tests and biochemical profiles on API 20 NE tests (bioMérieux, Marcy l'Étoile, France).⁸ All experiments were conducted in a BSL 3 laboratory.

MICs were determined by the agar dilution method. Antibiotic powders were obtained from their respective manufacturers and dissolved according to their recommendations. The antibiotics tested were: amoxicillin (Inava, France), co-amoxiclav (clavulanic acid 2 mg/L) (Smith-Kline-Beecham, France), piperacillin (Dakota, France), piperacillin/tazobactam (tazobactam 4 mg/L) (Wyeth-Lederle, France), ticarcillin and ticarcillin/clavulanate (Smith-Kline-Beecham, France), cefoxitin (Merck-Sharp-Dohme-Chibret, France), cefoperazone (Pfizer, France), cefsulodin (Takeda, France), cefotaxime (Roussel-Diamant, France), clavulanate (Smith-Kline-Beecham, France), tazobactam (Wyeth-Lederle, France), aztreonam (Sanofi-Winthrop, France), ceftazidime (Glaxo-Wellcome, France), imipenem (Merck-Sharp-Dohme-Chibret, France), nalidixic acid (Sanofi-Winthrop, France), pefloxacin (Bellon, France), ciprofloxacin (Bayer, France), ofloxacin (Roussel-Diamant, France), norfloxacin (Glaxo-Wellcome, France), gatifloxacin (Grünenthal, France), levofloxacin (Aventis, France), trimethoprim/sulfamethoxazole (co-trimoxazole; Roche, France), gentamicin (Schering-Plough), tobramycin (Lilly, France), netilmicin (Schering-Plough, France), amikacin (Bristol-Myers Squibb, France), chloramphenicol (Chauvin, France), fosfomycin (Sanofi-Winthrop, France), doxycycline (Asta-medica, France), minocycline (G-Gam, France), rifampicin (Roussel-Diamant, France), erythromycin (Abbot, France), clindamycin (Pharmacia-Upjohn, France) and novobiocin (Sigma). Bacteria were recovered prior to the experiment by placing the glycerol stock in Mueller–Hinton broth incubated at 37°C for 48 h. Colonies were obtained by plating 10 μL of the initial broth on blood agar medium incubated for 48 h at 37°C. After growth, each isolate was suspended in PBS and 100 μL of the suspension was added to 3.9 mL of Mueller–Hinton broth and incubated for 48 h in a water bath at 37°C. Mueller–Hinton agar plates were inoculated using a multiple Steer inoculator (Dynatech, UK) with a final inoculum of 10⁴ cfu/spot. The plates were incubated at 37°C to comply with the recommendations of the ‘Comité de l'Antibiogramme de la Société Française de Microbiologie’ (CASFM). These recommendations are freely available at http://www.sfm.asso.fr. Because B. mallei growth is usually poor after 24 h of incubation, the plates were incubated for 48 h and read at 24 and 48 h. Escherichia coli ATCC 25922 was used as a control strain. As there are no specific recommendations for the interpretation of susceptibility tests of B. mallei and B. pseudomallei, interpretative criteria were, when existing, those defined by the CASFM for B. cepacia (Table 2). When specific breakpoints were not available, general recommendations were applied.

Results and discussion

The results of susceptibility tests are shown in Table 2.

Whatever the antibiotic family, the present results demonstrated no differences between the strains isolated from humans and those isolated from animal or environmental sources.

Consistent with previous studies, the isolates tested in this protocol were highly resistant to amoxicillin, ticarcillin, cefoxitin, cefoperazone, cefsulodin and aztreonam. One strain, isolated from human infection, appeared resistant to ceftazidime with an MIC of 64 mg/L (MICs of ceftazidime for the other strains were between 1–4 mg/L). This strain also presented a cross-resistance to ticarcillin/clavulanate, doxycycline and minocycline and was categorized intermediate for co-amoxiclav with an MIC of 16 mg/L. All other strains of B. mallei and B. pseudomallei were susceptible to co-amoxiclav. Cefotaxime activity was low but was partially restored by clavulanate or tazobactam, demonstrating the effectiveness of β-lactamase inhibitors against the two species. The resistance profiles of both species for β-lactam antibiotics were similar; however, lower MICs were observed in B. mallei for β-lactam inhibitor combinations. MICs of piperacillin were low (0.125–8 mg/L) and those of the combination of piperacillin and tazobactam were three- or four-fold lower, suggesting that this antibiotic was hydrolysed by the chromosomal β-lactamases of both species.

All the isolates were susceptible to imipenem. This antibiotic was, with doxycycline and minocycline, one of the most active antibiotics tested. This has been observed previously in a study involving 211 clinical strains,⁵ and is of interest because this antibiotic is considered as a good alternative to ceftazidime in the treatment of disseminated disease. It has been recommended by the European Agency for the Evaluation of Medicinal Products (EMEA) for the treatment of suspected or confirmed melioidosis.⁹ The low MICs encountered with minocycline and doxycycline are also of interest because doxycycline has been used alone for the treatment of localized infection and has also been recommended by the EMEA in association with imipenem or meropenem for the treatment of severe cases of melioidosis.⁹ Despite the lack of recommendations, we assume that oral doxycycline could be useful for post-exposure prophylaxis.¹⁰

With 50% of isolates intermediate or resistant to ciprofloxacin (MIC breakpoint of 2 mg/L), this antibiotic cannot be recommended for treatment and/or prophylaxis. Clinical experience in maintenance therapy has demonstrated a poor efficacy for preventing relapses.¹¹ Both species demonstrated resistance to pefloxacin, ofloxacin and norfloxacin. MICs of gatifloxacin and levofloxacin were equivalent to those of ciprofloxacin for B. pseudomallei. All the strains of this species were resistant to erythromycin and clindamycin, and nearly all were resistant to all the aminoglycosides tested (gentamicin, tobramycin, netilmicin, amikacin). This resistance is due to the presence of a unique multidrug efflux system (AmrAB-OprA) in B. pseudomallei, which is specific for both aminoglycosides and macrolide antibiotics.¹² In contrast, MICs of aminoglycosides for B. mallei were lower and all strains appeared susceptible to netilmicin with MICs in the range 0.125–0.25 mg/L. All the MICs of clindamycin were high in both species, but only B. pseudomallei exhibited high MICs of erythromycin. This observation is consistent with those obtained with azithromycin.⁷B. pseudomallei was categorized as moderately susceptible to chloramphenicol. The MIC₅₀ and MIC₉₀ of co-trimoxazole were 8 and 16 mg/L, respectively, and the majority of the strains were categorized as intermediate or resistant (breakpoints ≤2/38, >8/152) to this antibiotic. This relative in vitro resistance is not correlated with clinical experience as co-trimoxazole has been traditionally used for the therapy of melioidosis. Such discrepancies between results obtained with co-trimoxazole by different susceptibility testing methods and clinical data have already been documented.¹

In conclusion, this study confirms the high level of antibiotic resistance in B. mallei and B. pseudomallei, including towards agents not tested in previous studies. Nevertheless, the overall resistance is lower in B. mallei. The resistance profiles appear to be independent of the origin of isolates. Imipenem, ceftazidime, co-amoxiclav, piperacillin, piperacillin/tazobactam and doxycycline appear as the more effective drugs tested on this panel of isolates. These results remain consistent with the current recommendations for the treatment of melioidosis and glanders. However, the emergence of ceftazidime-resistant clinical isolates and the wide distribution of B. pseudomallei in Southeast Asia increase the risk of malicious use of those resistant strains. Piperacillin/tazobactam could be a useful alternative for treatment of both glanders and melioidosis. Nevertheless, the disparity between in vitro results and clinical response underlines the necessity to validate this association by time–kill studies, animal models and clinical experience.

Table 1.

Strains used in this study

Organism	Strain reference	Isolated from	Year of isolation	Country
B. mallei	ATCC 23344 (T)	man	1966	China
B. mallei	NCTC 10248	man	1950	Turkey
B. mallei	NCTC 10260	man	1949	Turkey
B. mallei	ATCC 10399	horse	1948	China
B. mallei	CIP 64.12	horse	1961	Hungary
B. mallei	CIP A.198	horse	1948	Iran
B. mallei	CIP A.200	horse	1948	Iran
B. mallei	NCTC 03709	horse	1932	India
B. mallei	NCTC 10230	horse	1961	Hungary
B. mallei	CIP 52.236	mule	1952	Iran
B. mallei	NCTC 03708	mule	1931	Mule
B. mallei	CIP A.187	NA	NA	NA
B. mallei	NCTC 120 A	NA	1920	NA
B. mallei	NCTC 10229	NA	1961	Hungary
B. mallei	NCTC 10247	NA	1960	Turkey
B. pseudomallei	ATCC 23343 (T)	man	1954	NA
B. pseudomallei	CIP 52.238	man	1952	Vietnam
B. pseudomallei	CIP 52.239	man	1952	Vietnam
B. pseudomallei	CIP 55.135	man	1955	Vietnam
B. pseudomallei	CIP 60.67	man	1953	Vietnam
B. pseudomallei	CIP 60.68	man	1948	Vietnam
B. pseudomallei	CIP 60.68 (attenuated variant)	man
B. pseudomallei	CIP 62.27	man	1962	NA
B. pseudomallei	CIP 62.28	man	1962	France
B. pseudomallei	CIP A.202	man	1947	Vietnam
B. pseudomallei	CIP A.203	man	1947	Vietnam
B. pseudomallei	CRSSA 047/98	man	1998	France
B. pseudomallei	CRSSA 008/93	man	1993	France
B. pseudomallei	CRSSA 041/97	man	1997	France
B. pseudomallei	CRSSA 042/97	man	1997	France
B. pseudomallei	CRSSA 043/97	man	1997	France
B. pseudomallei	CRSSA 050/98	man	1998	France
B. pseudomallei	CRSSA 062/00	man	2000	France
B. pseudomallei	CRSSA 067/00	man	2000	France
B. pseudomallei	IPP 70061	man	NA	Vietnam
B. pseudomallei	IPP Ducruet	man	1963	France
B. pseudomallei	IPP PA	man	NA	Australia
B. pseudomallei	NCTC 6700	man	1944	NA
B. pseudomallei	NCTC 10274	man	1961	Malaya
B. pseudomallei	NCTC 10276	man	1962	NA
B. pseudomallei	PHLS SID 3871	man	1999	England
B. pseudomallei	PHLS SID 3783	man	1999	England
B. pseudomallei	PHLS SID 4717	man	1999	England
B. pseudomallei	PHLS SID 4718	man	1999	England
B. pseudomallei	PHLS SID 4935	man	2000	England
B. pseudomallei	PHLS SID 5278	man	1966	England
B. pseudomallei	IPP 77804	dolphin	NA	Honk-Kong
B. pseudomallei	IPP 56.91	goat	NA	Chad
B. pseudomallei	CRSSA 089/01	lizard	2001	France
B. pseudomallei	ATCC 15682	monkey	1961	NA
B. pseudomallei	NCTC 04846	monkey	1935	Singapore
B. pseudomallei	IPP A120	ovine	NA	Vietnam
B. pseudomallei	CIP 68.2	pig	1967	Malaya
B. pseudomallei	CIP 68.3	pig	1967	Malaya
B. pseudomallei	NCTC 1688	rat	1923	Malaya
B. pseudomallei	NCTC 8016	sheep	1949	Australia
B. pseudomallei	PHLS E090	soil	1990	Thailand
B. pseudomallei	IPP NT16	soil	NA	NA
B. pseudomallei	IPP W5	water	NA	NA
B. pseudomallei	ATCC 11668	NA	NA	China
B. pseudomallei	NCTC 7383	NA	1948	NA
B. pseudomallei	NCTC 7431	NA	1948	NA
B. pseudomallei	NCTC 8707	NA	1946	Singapore
B. pseudomallei	NCTC 8708	NA	1946	Singapore
B. pseudomallei	PHLS E222	NA	NA	NA

ATCC, American Type Culture Collection (Manassas, VA, USA); NCTC, National Collection of Type Cultures (London, UK); CIP, Collection de l'Institut Pasteur (Paris, France); CRSSA, Centre de Recherches du Service de Santé des Armées (La Tronche, France); IPP, strains kindly provided by Professor A. Dodin (Institut Pasteur, Paris, France; now deceased); PHLS, strains kindly provided by Dr T. Pitt (Public Health Laboratory Service, London, UK); NA, not available.

Table 2.

MIC₅₀s, MIC₉₀s and ranges for 50 B. pseudomallei and 15 B. mallei strains

		B. pseudomallei (n=50)							B. mallei (n=15)
	Breakpoint for resistance	MIC50	MIC90	range	percentage resistant	MIC50	MIC90	range	percentage resistant
Amoxicillin	>16a	64	>128	16–128	98	64	>64	16–>128	87
Co-amoxiclav	>16/2a	8	8	0.5–64	2	0.125	4	0.125–8	0
Ticarcillin	>64b	64	>128	16–>128	100	64	>128	32–>128	100
Ticarcillin/  clavulanate	>64/4b	8	16	0.5–128	8	4	4	0.125–8	0
Piperacillin	>64b	4	8	0.25–8	0	4	8	0.125–8	0
Piperacillin/  tazobactam	>64/4b	0.25	1	0.25–4	0	0.25	0.25	0.125–8	0
Cefoxitin	>32a	>128	>128	>128	100	>128	>128	4–>128	87
Cefoperazone	>32a	16	16	8–>128	14	8	16	4–64	6
Cefsulodin	>32b	>128	>128	>128	72	>128	>128	32–>128	87
Cefotaxime	>32b	16	16	1–64	4	4	16	0.5–32	0
Cefotaxime/  clavulanate	ND	8	8	0.06–32	ND	4	32	0.5–32	ND
Cefotaxime/  tazobactam	ND	4	8	0.06–32	ND	2	8	0.06–8	ND
Aztreonam	>32b	32	64	8–64	16	32	64	4–128	25
Ceftazidime	>32b	2	4	1–64	2	2	2	1–4	0
Imipenem	>8b	1	1	0.5–1	0	0.25	0.5	0.125–0.5	0
Nalidixic acid	>16b	16	32	8–>128	52	16	32	8–32	25
Pefloxacin	>4b	8	16	1–>128	72	2	4	0.125–16	12
Ciprofloxacin	>2b	2	4	0.25–32	50	2	4	0.5–16	12
Ofloxacin	>4b	4	8	1–>128	20	2	2	0.125–32	6
Norfloxacin	>2a	32	64	8–>128	100	8	16	8–128	100
Gatifloxacin	>2b	2	4	0.5–32	48	0.5	1	0.125–1	12
Levofloxacin	>4a	2	4	1–32	52	1	1	0.125–4	0
Gentamicin	>8b	32	64	0.25–>128	98	2	128	0.125–128	19
Tobramycin	>8a	128	>128	2–>128	94	2	4	0.125–16	12
Netilmicin	>8b	128	>128	4–>128	98	0.125	0.125	0.125–0.25	0
Amikacin	>16b	64	128	4–128	92	32	64	1–128	70
Co-trimoxazole	>8/152b	8/152	16/304	1/19–64/1216	68	2	4	1–4	0
Chloramphenicol	>16b	8	16	2–16	4	2	4	0.125–8	0
Fosfomycin	>32a	>128	>128	>128	100	>128	>128	>128	100
Doxycycline	>8b	1	1	0.125–1	2	0.125	0.25	0.125–0.5	0
Minocycline	>8b	1	2	0.125–1	2	0.125	0.25	0.25–1	0
Rifampicin	>16b	16	32	0.25–128	88	2	4	0.25–16	0
Erythromycin	>4b	>64	>64	8–>64	100	0.25	1	0.25–2	0
Clindamycin	>2b	>128	>128	>128	100	>128	>128	>128	100
Novobiocin	ND	8	8	0.25–16	ND	2	4	0.25–4	ND

a

Breakpoints defined in the general recommendations of the ‘Comité de l'Antibiogramme de la Société Française de Microbiologie’ (CASFM).

b

Breakpoints defined for B. cepacia by the CASFM.

ND, no interpretative criteria.

We are indebted to Isabelle Perrichon for excellent technical work. This work was supported by the French Ministry of Defense, grant no. 3-e/LR/EMA.

Author notes

1Centre de Recherches du Service de Santé des Armées Emile Pardé, Département de Biologie des Agents Transmissibles, 24 Avenue des Maquis du Grésivaudan, B.P. 87, F-38702 La Tronche; 2Service de Biologie Médicale, Hôpital d'Instruction des Armées Begin, 69 Avenue de Paris, F-94160 Saint-Mandé, France