Effects of a combination of amlodipine and imipenem on 42 clinical isolates of Acinetobacter baumannii obtained from a teaching hospital in Guangzhou, China

Acinetobacter baumannii (AB) has emerged as one of the most troublesome pathogens in health care institutions globally. It is characterised by frequent episodes of multi-drug resistance due to multiple mechanisms, and therapeutic options have become increasingly restricted [1‐3]. Unfortunately, the present rate of discovery of antibiotics is much slower than that in the past [4]. This problem is exemplified by the fact that it is necessary to identify potential reservoirs of the organism and modes of transmission to control the spread of A. baumannii in hospitals. To distinguish the outbreak strain from epidemiologically unrelated acinetobacters, a comparison of isolates at the subspecies level is required. Multilocus sequence typing (MLST) is a highly discriminative typing system in which typing data are translated into a numerical code that can be obtained in an identical manner at different laboratories using the same protocol. It provides a portable method that may be suitable for global epidemiologic studies and that may allow for the recognition of epidemic, multiresistant, and virulent A. baumannii clones and the monitoring of their national and international spread [1, 5, 6]. However, data have been limited in Guangzhou.

Recently, an effective approach has been to explore non-antibiotic compounds (helper compounds) that express antibacterial properties, possibly by acting through different mechanisms from those of existing drugs, either by the enhancement of antibiotic activity (synergism) or the reversal of antibiotic resistance, yielding sensitivity comparable to that derived from classical antibiotics in previously drug-resistant microorganisms [7]. Among these non-antibiotics, amlodipine (AML) is the most promising helper compound. It has been shown to exhibit in vitro activity against a wide range of bacteria [8‐11], but few studies have reported its potentiating effects on clinical A. baumannii isolates. In this study, AML was tested, alone and in combination with imipenem, against A. baumannii isolates.

A. baumannii infections have become more difficult to treat due to the emergence of isolates that are resistant to multiple antimicrobial drugs. Data from the Chinese Network for Bacterial Resistance Surveillance demonstrated that almost 17% of these isolates were resistant to all of the antimicrobial agents that are routinely used in clinical practice and that the susceptibility of A. baumannii to carbapenems has decreased to less than 50% [16]. As a consequence, infections caused by A. baumannii strains have been associated with high mortality and treatment failure [17]. According to our data, A. baumannii has been widely distributed in our hospital and has been isolated from the respiratory system, blood, wounds, urine, and cerebrospinal fluid, especially from ICU patients who are older and have longer inpatient stays, similar to the results reported in many other studies [1, 18‐20]. The susceptibility of these isolates to imipenem was 16.1% (according to the MIC method). Only 1 (2.4%) MBL-producing A. baumannii strain was found in our study, which was similar to the findings of Hui Wang [21], in which only 1 isolate among 221 nonrepetitive imipenem-resistant clinical isolates of Acinetobacter spp. carried the bla(IMP-8) MBL gene. The present study demonstrated that MBL was not the primary cause of A. baumannii resistance to imipenem in our hospital. Class D β-lactamases, the modification of penicillin-binding proteins and porins, or the upregulation of the AdeABC efflux system may have played more important roles in their potential resistance mechanisms [22]; thus, further investigation is needed.

MLST is an unambiguous typing method that has achieved notable success in global epidemiological investigations of A. baumannii[5, 6]. Recently, some studies [6, 23‐26] have revealed that CC92 (previously known as CC22) has worldwide dissemination. The ST92 A. baumannii isolate is the most common clone, and it has been predicted to be the founder of this clonal complex. Yiqi Fu et al. [26] confirmed that CC92 represented the most epidemic STs in China. In addition to ST92, the other STs belong to CC92 and vary by area. A study by He et al. [25] showed that more ST138 isolates are associated with ST92 in western China. However, results from Qiao Zhong et al. [23] indicated that ST75, accompanied by ST92, may be the most common epidemical sequence types in eastern China. In our study, we assumed that ST195 and ST208 may be more common in southern China. These differences may be caused by different antibiotic usage habits, which could have influenced the evolutionary direction of ST92. Why has this clonal complex been so successful? Studies have shown that antibiotic susceptibility of CC92 is variable [27]. These findings may suggest that adaptation to the hospital environment, as well as antibiotic resistance, has been more important for the success of A. baumannii as a nosocomial pathogen.

Because imipenem remains the first-choice agent in China against A. baumannii strains [28], finding a way to enhance the antibiotic activity (synergism) of carbapenems against A. baumannii strains could be encouraging. Calcium channel blockers (CCBs) are a heterogeneous class of drugs used for the control of hypertension, angina, and cardiac arrhythmias [29]. However, many studies have reported a wide variety of broad-spectrum antibacterial activities of CCBs [30], the most powerful of which was AML. Kumar [9] reported in vitro antimicrobial activity of AML against several Gram-positive and Gram-negative standard and clinical strains, including Bacillus pumilus NCTC 8241, Staphylococcus aureus NCTC 6571, Escherichia coli K12Row, Salmonella typhimurium NCTC74, Shigella dysenteriae 7 NCTC 519/66, Shigella sonnei 1NCTC 5/59, Shigella flexneri 4a24, Shigella boydii 8 NCTC 254/66, Klebsiella pneumoniae 14, Vibrio cholerae ATCC 14033, and P. aeruginosa APC. The inhibitory concentrations of AML against most of the strains tested ranged from 25 to 200 μg/ml, but its effects on P. aeruginosa, E. coli, and K. pneumoniae were limited, and the inhibitory concentrations were >800, >800, and 400 μg/ml, respectively. A further study demonstrated that AML in combination with streptomycin had a synergistic effect [10]. In addition to its in vitro and in vivo activity against Gram-positive and Gram-negative bacteria, AML showed a two- to eightfold reduction in its MIC when combined with streptomycin, and vice versa [10]. Similarly, when paired with levofloxacin, AML showed potential synergistic effects in the eradication of P. aeruginosa biofilms [11]. Our research demonstrated the antimicrobial activities of AML against clinical A. baumannii strains, and the MICs ranged from 40 to 320 μg/ml. In combination with imipenem, half of the isolates (N = 21, 50.0%) demonstrated synergy or partial synergy. Further investigations, such as the development of a time-killing curve, which may demonstrate killing synergistic effects against A. baumannii strains, are needed.

To date, research on the antimicrobial efficacy of AML has been conducted only in vitro and in animal models. Chronic toxicity tests with healthy mice showed that the therapeutic dose of AML (10 mg/kg) was within a safe and acceptable margin. Low doses of AML at 1.5 mg/kg/day continue to show good efficacy in mice [7]. According to our study results, AML must be used at doses of up to 40 μg/ml to improve the activity of imipenem against A. baumannii isolates. However, in humans, the recommended dose of AML is much lower, at only 10 mg/day (Cmax 6.14 ng/ml) as an antihypertensive agent, thus currently limiting the clinical application of AML [31]. However, AML has been able to cure highly virulent bacterial infections in mice. This ability suggests that AML could be employed as a ‘lead compound’ to synthesise more active novel agents that may be free of side effects, such as hypotension. Because both drugs have been satisfactorily used for a long time in clinical medicine for different purposes, their combination in humans is feasible.