The In Vitro Antibiotic Susceptibility of Malaysian Isolates of Burkholderia pseudomallei

Acute melioidosis may present as localised or septicaemic infections and can be fatal if left untreated. Burkholderia pseudomallei resistant to antibiotics used for the treatment of melioidosis had been reported. The aim of this study was to determine the in vitro antibiotic susceptibility patterns of Burkholderia pseudomallei isolated in Malaysia to a panel of antibiotics used for the treatment of melioidosis and also to potential alternative antibiotics such as tigecycline, ampicillin/sulbactam, and piperacillin/tazobactam. A total of 170 Burkholderia pseudomallei isolates were subjected to minimum inhibitory concentration determination using E-test method to eleven antibiotics. All isolates were sensitive to meropenem and piperacillin/tazobactam. For ceftazidime, imipenem, amoxicillin/clavulanic acid, and doxycycline resistance was observed in 1 isolate (0.6%) for each of the antibiotics. Trimethoprim/sulfamethoxazole resistance was observed in 17 (10%) isolates. For other antibiotics, ampicillin/sulbactam, chloramphenicol, tigecycline, and ciprofloxacin resistance were observed in 1 (0.6%), 6 (3.5%), 60 (35.3%) and 98 (57.7%) isolates respectively. One isolate B170/06 exhibited resistance to 4 antibiotics, namely, ciprofloxacin, chloramphenicol, trimethoprim/sulfamethoxazole, and tigecycline. In conclusion, the Malaysian isolates were highly susceptible to the current antibiotics used in the treatment of melioidosis in Malaysia. Multiple resistances to the antibiotics used in the maintenance therapy are the cause for a concern.


Introduction
The causative agent of melioidosis, Burkholderia pseudomallei, is endemic in the Northern part of Australia and Southeast Asia including Malaysia. Acute melioidosis may present as localized or septicaemic infections and can be fatal if left untreated. It was the common cause of community-acquired pneumonia in Northeastern Thailand and was attributed as the cause of fatal community-acquired bacteremic pneumonia in Northern Australia [1,2]. Latent infection may remain asymptomatic for years only to be reactivated from a latent focus when the host is immunocompromised. Therefore, it is important to treat melioidosis with prolonged course of antibiotics so as to avoid disease relapses which are commonly associated with short courses of antibiotics. Some patients may default treatment or take improper dosage of antibiotics because of the long duration of treatment, and this may contribute to the relapse or the development of resistance.
Burkholderia pseudomallei is intrinsically resistant to a wide range of antibiotics which include some -lactam antibiotics, aminoglycosides, and macrolides. The antibiotics that are currently being used for the therapy of melioidosis are ceftazidime, imipenem, meropenem, amoxicillin/clavulanate, cefoperazone/sulbactam, trimethoprim-sulfamethoxazole (TMP-SMX), doxycycline, and chloramphenicol. The antibiotic regime for the treatment of melioidosis varies from one country to another. In Malaysia, ceftazidime alone, or in combination with TMP-SMX or cefoperazone/sulbactam alone or in combination with TMP-SMX or imipenem, is the recommended antibiotic of choice for the intensive phase of treatment followed by oral TMP-SMX plus doxycycline or amoxicillin/clavulanate in the maintenance phase [3]. The development of resistance of Burkholderia pseudomallei to some of these antibiotics has been reported in neighbouring countries such as Singapore and Thailand [4,5]. Reports on the antibiotic susceptibility of Burkholderia pseudomallei in 2 International Journal of Microbiology Malaysia have been limited to a few selected antibiotics and a smaller number of tested strains and from restricted demographic areas [6,7]. Therefore, the aim of the study was to determine the in vitro antibiotic susceptibility patterns of clinical isolates of Burkholderia pseudomallei to a panel of antibiotics used for the treatment of melioidosis and also to the potential alternative antibiotics such as tigecycline, ampicillin/sulbactam, and piperacillin/tazobactam.  (Table 1). These isolates were sent from these hospitals for the confirmation of identification at the Bacteriology Unit, Institute for Medical Research. Species identification was carried out by Gram-staining, motility, API 20NE (bioMèrieux), and polymerase chain reaction technique using specific 16rRNA primers as described by Brook et al. 1997 [8]. The strains were stored at −80 ∘ C in 20% glycerol and were revived by subculturing onto blood agar plates before being further used.

Results
All strains were sensitive to meropenem and piperacillin/tazobactam. Sensitivity to ceftazidime, imipenem, amoxicillin/clavulanic acid, ampicillin/sulbactam, and doxycycline was noted in 169 (99.4)% of the isolates. One isolate was shown to have heterogenous population with ceftazidime susceptibility of 8 g/mL and ceftazidime resistance of ≥256 g/mL. This strain, however, remained susceptible to other antibiotics. The isolate that was intermediately resistant to imipenem (MIC 6 g/mL) was also resistant to amoxicillin/clavulanic acid and ampicillin/sulbactam. For chloramphenicol and TMP-SMX, 164 (96.5%) and 153 (90%) of the strains were susceptible, respectively. Only 72 (42.3%) isolates were susceptible to ciprofloxacin while 94 (55.3%) isolates showed intermediate resistance with MIC of 1.5-3.0 g/mL. Susceptibility to tigecycline was observed in 110 isolates (64.7%) while intermediate resistance was noted in 59 (34.7%) of the isolates. One isolate was resistant to tigecycline at the MIC of 8.0 g/mL.
The minimum concentration of antibiotics that inhibited 50% (MIC 50 ) and 90% (MIC 90 ) of the isolates and the range of MICs of the tested antibiotics were shown in Table 2. The MIC 90 of most of the antibiotics were within the range of 0.38 to 4.0 g/mL. Chloramphenicol was shown to have higher MIC 50 and MIC 90 values than the other antibiotics (MIC 50 , 6.0 g/mL; MIC 90 8.0 g/mL). Imipenem (MIC 50, 0.38 g/mL; MIC 90 , 0.38 g/mL) was more active towards these strains than meropenem (MIC 50 , 0.75 g/mL and MIC 90 , 1.0 g/mL).
The resistance pattern observed among these strains was mainly monoresistance where 7 strains showed resistance to a single antibiotic, either to ceftazidime (1 strain), ciprofloxacin (1 strain), or trimethoprim/sulfamethoxazole (5 strains). Two strains showed resistance to 2 different types of antibiotics (1 isolate to ciprofloxacin & trimethoprim/sulfamethoxazole; 1 strain to amoxicillin/clavulanic acid and ampicillin/sulbactam). One strain showed multiple resistances to 4 antibiotics namely ciprofloxacin, chloramphenicol, trimethoprim/sulfamethoxazole, and tigecycline and intermediate resistance to doxycycline.

Discussion
This study showed that the local strains were highly susceptible to meropenem and piperacillin/tazobactam. For ceftazidime, imipenem, amoxicillin/clavulanic acid, and doxycycline, only 0.6% of the isolate was resistant to these antibiotics. A highly resistant subpopulation of ceftazidime resistance was detected in one of the strain. This was a blood isolate from a patient who had no past record of melioidosis and had passed away a day after admission before the culture result was obtained. Primary resistance to ceftazidime is rare and studies had documented that the development of resistance emerge mainly during treatment [11,12].
The susceptibility to trimethoprim/sulfamethoxazole is slightly lower compared to 97.5% in the another local study on 80 Burkholderia pseudomallei [6]. Only 4.0% of our isolates was resistant to trimethoprim/sulfamethoxazole compared    to 13% resistance rate in Thailand [13]. Our local strains were also highly susceptible to chloramphenicol. An open label study in Thailand had shown that the combination of chloramphenicol, TMP-SMX, and doxycycline was associated with higher adverse effect compared to TMP-SMX and doxycycline only [14]. Chloramphenicol was not included in the maintenance therapy of melioidosis in Malaysia [3]. The antibiotic susceptibility rate of ciprofloxacin was low at 42.3%, and 90% of the isolates were inhibited at the intermediate MIC value of 2 g/mL, which implied that ciprofloxacin was less effective towards these strains. Fluoroquinolones have been shown to be less effective clinically [15,16].
A study on the activity of tigecycline against Burkholderia pseudomallei in Thailand showed low MIC 50 and MIC 90 (0.5 g/mL and 2.0 g/mL) [17]. This finding is similar to another study in Malaysia [7]. In contrast, the isolates tested in this study were less susceptible to tigecycline, where higher MIC 50 and MIC 90 values (0.75 g/mL and 4.0 g/mL) were observed. The susceptibility of tigecycline was only 64.7%, and 34.7% of the strains were inhibited at the intermediate range of MIC of 3.0-6.0 g/mL. This is in concordance with another study in Australia, where 85.5% of the isolates were inhibited at an intermediate MIC of 4.0 g/mL with lower susceptibility rate of 14.5% [18].