The Distribution of Carbapenem-Resistant Acinetobacter Species and High Prevalence of CC92 OXA-23-Producing Acinetobacter Baumannii in Community Hospitals in South Korea

Introduction

Acinetobacter species, which are aerobic non-fermentative gram-negative bacilli, are leading primary nosocomial pathogens both in hospital-acquired infections and community-acquired infections.^1,2 To date, Acinetobacter isolates are divided into 144 different Acinetobacter species, including Acinetobacter baumannii, Acinetobacter nosocomialis, and Acinetobacter pittii, which are the most frequently identified Acinetobacter species, and Acinetobacter tianfuensis species and Acinetobacter rongchengensis species, which are the Acinetobacter species that were most recently identified.^3,4 These organisms often cause respiratory infections, urinary tract infections, meningitis endocarditis, bloodstream infections and burn infections.⁵

Carbapenems are used as one of the first-line choices for treatment due to the extensive antimicrobial resistance in Acinetobacter isolates in hospitals. However, when β-lactamases, which can hydrolyze β-lactam antibiotics including carbapenems, are acquired, the treatment of Acinetobacter infections is difficult because of the acquired carbapenemases.⁶ The mechanism of resistance to carbapenems in A. baumannii strains has been primarily mediated by the production of OXA-type class D carbapenemases, especially the bla_OXA-23 gene associated with insertion sequences (IS), primarily by ISAba1, indicating generally ISAba1-bla_OXA-23 structures.⁷ In non-baumannii Acinetobacter species, class A and B carbapenemases, including a variety of GES-type class A carbapenemases and IMP-, VIM-, and NDM-type metallo-β-lactamases of class B carbapenemases, have been frequently detected.⁸

The outbreak of A. baumannii strains harboring bla_OXA-23 has been mainly attributed to clonal complex 92 (CC92) or global clone 2 (GC2) in the Oxford scheme worldwide,⁹ which means these clones are resistant to carbapenem as major clinical pathogens. Same here in South Korea, the various infections by CC92 A. baumannii clinical strains has difficulty due to limited therapeutic options in clinical field.^6,10,11 In South Korea, the prevalence and epidemiology of carbapenem-resistant Acinetobacter species have been well reported as hospital-associated pathogens originating from tertiary hospitals and university hospitals in Korea global antimicrobial resistance (AMR) surveillance systems,^10,11 while few reports on the characteristics of carbapenem-resistant Acinetobacter species originating from community hospitals have been published.

Here, we describe the prevalence and distribution of carbapenem-resistant Acinetobacter species isolates originating from community hospitals, including long-term care facilities and general hospitals, to study nationwide AMR in South Korea.

Materials and Methods

Bacterial Collection and Identification

In this study, we set that above Acinetobacter species isolated from two patient groups would have caused community-acquired infections for the following reasons. I) Patient group who have not recently been to a tertiary hospital and are admitted to a long-term care hospital, and II) patient group who have not recently been to a tertiary hospital and have received outpatient care in a local general hospital. All specimens originating from two patient groups were only collected as part of the surveillance of Acinetobacter species.

Acinetobacter species were cultured on MacConkey agar (Duksan Science, Seoul, Korea). The identities of the isolates were confirmed by bla_OXA-51 PCR to be A. baumannii and were identified by rpoB gene sequence analysis as non-baumannii Acinetobacter species.

A total of 817 non-duplicated Acinetobacter clinical isolates were obtained from 817 clinical specimens of patients suspected respiratory infections, urinary tract infections, and bloodstream infections between December 2022 and July 2023 at long-term care facilities and general hospitals in 16 regions geographically distributed throughout South Korea. We only included the strains identified as Acinetobacter species using OXA-51 and/or rpoB gene analysis in this study in case positively cultured on MacConkey agar.

Antimicrobial Susceptibility Testing

Acinetobacter isolates were evaluated for ampicillin/sulbactam, piperacillin, piperacillin/tazobactam, cefotaxime, ceftazidime, cefepime, aztreonam, imipenem, meropenem, amikacin, gentamicin, ciprofloxacin, minocycline, tigecycline, cotrimoxazole, and ticarcillin/clavulanic acid susceptibility via VITEK-2 (bioMérieux, Inc., Durham, NC). The quality control strains used were Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853. The minimal inhibitory concentrations (MICs) of imipenem and meropenem were determined by broth microdilution according to the guidelines of the Clinical and Laboratory Standards Institute (CLSI). The interpretations were performed according to CLSI guidelines (CLSI M100-S32, 2023).

PCR Assays and Sequencing

Detection of the carbapenemase genes (bla_OXA-23-like, bla_OXA-48-like, bla_OXA-58-like, bla_IMP-type, bla_VIM-type, bla_NDM-type, bla_GES-type, and bla_KPC-type) was carried out in carbapenem-resistant isolates using conventional single PCR and sequence analysis under the following amplification conditions: 94 °C for 5 min; 35 cycles of 94 °C for 30 sec, 55–60 °C for 30 sec and 72 °C for 30 sec; and a final extension at 72 °C for 5 min using a C1000TM Thermal Cycler (Bio-Rad, Hercules, CA). The sequences were compared to published DNA sequences using BLAST. The primers used in this study are summarized in Table S1.

Epidemiological Typing

The epidemiological relationships among the bla_OXA-23-producing A. baumannii isolates were analyzed by multilocus sequence typing (MLST). PCR amplification and sequencing of partial fragments of seven housekeeping genes (gltA, gyrB, gdhB, recA, cpn60, gpi and rpoD) were performed according to the Oxford scheme, and the experimentally determined amplicons of both strands were compared to preexisting sequences in the MLST database to assign allelic numbers and sequence types (STs) (https://pubmlst.org/organisms/acinetobacter-baumannii).¹² Clonal complexes (CCs) are defined as groups of STs with similar allelic profiles; every ST shares at least five or more of the seven alleles that are involved in predicting ancestral epidemiological genotype using the BURST algorithm.¹³ In this regard, five genes (gltA, gdhB, recA, cpn60, and rpoD) were amplified to assign CC92 (the allelic profile is gltA-1, gdhB-3, recA-2, cpn60-2, and rpoD-3) in this study. If different allele numbers were obtained, the gyrB and gpi genes were amplified to complete the ST.

Ethics Statement and Patients Consent

The Ethics Committee of Silla University approved the study. All patients were included after providing written informed consent and/or assent as appropriate. This research did not involve an interventional study. For personal protection, we were only provided with anonymized data including isolated sample type and patients’ address. Therefore, this study was conducted following the Declaration of Helsinki.

Results

The prevalence of Acinetobacter clinical isolates identified from each of the 16 regions geographically distributed throughout South Korea is shown in Figure 1. A total of 817 non-duplicated Acinetobacter species isolates were collected, consisting of 659 A. baumannii harboring blaOXA-51 variants, 55 A. pittii, 20 Acinetobacter bereziniae, 19 A. nosocomialis, 19 Acinetobacter johnsonii, 16 Acinetobacter lwoffii, 5 Acinetobacter junii, 4 Acinetobacter ursingii, 4 Acinetobacter seifertii, 4 Acinetobacter soli, 3 Acinetobacter guillouiae, 2 Acinetobacter radioresistens and one Acinetobacter baylyi, Acinetobacter colistiniresistens, Acinetobacter courvalinii, Acinetobacter dispersus, Acinetobacter haemolyticus, Acinetobacter oleivorans and Acinetobacter vivianii. The list of above 158 non-baumannii Acinetobacter species isolates from each of the 16 regions is shown in Table S2.

Figure 1 Map showing the prevalence and characteristics of Acinetobacter clinical isolates in 16 regions in South Korea. The shade of Orange indicates the percentage each of carbapenem-resistant Acinetobacter species in this study. The gray and dark gray colors in the circle indicate the number of non-baumannii Acinetobacter and A. baumannii isolates, respectively. The white numbers in each colored pentagon arrow indicate the carbapenem resistance determinants.

Among the all 817 samples positively cultured from various clinical specimens on MacConkey agar, the predominant sample type was sputum (n=439, 53.7%), urine (n=195, 23.9%), and wound/pus (n=117, 14.3%), followed by other sample types (n=23), blood samples (n=17), throat swabs (n=9), vaginal discharge (n=5), nasal swabs (n=4), catheter tips (n=3), ear discharge (n=3), and body fluids (n=2).

The antimicrobial susceptibility profiles of the A. baumannii and non-baumannii Acinetobacter isolates are shown in Figure 2. The resistance rate to carbapenems (imipenem and meropenem) was 87.4% (n=576) in the A. baumannii clinical isolates. Among the non-baumannii Acinetobacter isolates, 8.9% were carbapenem-resistant (n=14). The level of carbapenem resistance (R%) in 16 regions of South Korea represent in Figure 1.

Figure 2 Antimicrobial susceptibility of Acinetobacter clinical isolates in this study. (A) A. baumannii isolates (n=659); (B) Non-baumannii Acinetobacter species (n=158). Abbreviations: S, susceptible; I, intermediate; R, resistant; SAM, Ampicillin/sulbactam; PIP, piperacillin; TZP, Piperacillin/tazobactam; CTX, cefotaxime; CAZ, ceftazidime; FEP, cefepime; ATM, aztreonam; IPM, imipenem; MEM, meropenem; AMK, amikacin; GEN, gentamicin; CIP, ciprofloxacin; MIN, minocycline; TGC, tigecycline; TIM, Ticarcillin/Clavulanic Acid.

Above 590 carbapenem-resistant Acinetobacter isolates also exhibited prominently high rates (ca. >90%) of resistance to piperacillin, piperacillin/tazobactam, cefotaxime, ceftazidime, cefepime, aztreonam, gentamicin, ciprofloxacin, trimethoprim/sulfamethoxazole, and ticarcillin/clavulanic acid, in contrast to the resistance rates to tigecycline (43.2% 255/590), amikacin (16.3%, 96/590), and minocycline (10.2% 60/590).

The main mechanism of resistance to carbapenems in A. baumannii isolates is the production of oxacillinases encoded by bla_OXA-23-like without the class A or B carbapenemases, while the production of bla_NDM-1 (n=3) and bla_VIM-2 (n=2) with bla_OXA-23 (n=9) occurred in 14 non-baumannii Acinetobacter species. The Information of the 14 carbapenemase-producing non-baumannii Acinetobacter isolates is presented in Table 1.

Table 1 The Profiles of 14 Carbapenemase-Producing Non-Baumannii Acinetobacter Isolates

The imipenem and meropenem MIC ranges of the 576 carbapenem-resistant A. baumannii (CRAB) isolates were 32 - >256 µg/mL for both antibiotics. The carbapenem MIC range of the 14 carbapenem-resistant non-baumannii Acinetobacter isolates was 16–128 µg/mL for imipenem and 16–64 µg/mL for meropenem (Table 1). The MIC₅₀ and MIC₉₀ of carbapenems in the 576 CRAB isolates were 64 µg/mL and 128 µg/mL for imipenem and meropenem, respectively.

MLST typing was performed on all 576 CRAB isolates harboring bla_OXA-23. The results of MLST analysis revealed that the genotype of the CRAB isolates was relatively simple. A total of two STs (ST373 and ST447) were identified as singletons, and four CCs (CC20, CC92, C108, and CC110) were identified. MLST analysis revealed that CC92 or global clone 2 (GC2) (n=519, 90.1%) was the most prevalent clone, and this clone was found in every region. Furthermore, ST229 belonging to CC110 was identified (n=47, 8.2%), followed by ST373 (n=7) and ST447 (n=1) as a singleton, one ST397 belonging to CC108 or GC1. and one ST620 belonging to CC20 [Table 2].

Table 2 The Characteristics of A. Baumannii Related to Sequence Types (STs) and Clonal Complexes (CCs) Identified in This Study

Discussion

Acinetobacter species are known as opportunistic pathogens that commonly cause nosocomial infections in hospitalized patients.¹⁴ Because Acinetobacter clinical isolates generally exhibit multidrug resistance pattern, carbapenems have been regarded as effective treatment options for Acinetobacter infections in hospitals. However, the increasing use of carbapenems has resulted in the prevalent emergence of carbapenem-resistant Acinetobacter species, especially CRAB isolates.^6,15 After many reports indicating the widely spread of CRAB across several provinces in China,^16,17 CRAB outbreaks have been regarded as a global nosocomial menace.^9,18 In most cases, they can become resistant to carbapenems through the production of the bla_OXA-23 gene, which is the major mechanism involved in the acquisition of carbapenem resistance.³ According to reports from South Korea global AMR surveillance systems,^10,11 the rates of resistance to carbapenems mostly associated with bla_OXA-23 were 89.9% in 2016 and 90.5% in 2017–2019, for A. baumannii isolates originating from tertiary hospitals and university hospitals. In this study, 87.4% of the A. baumannii isolates originating from community hospitals were resistant to carbapenems. Based on these results, CRAB has already become a menacing pathogen in South Korea, regardless of whether the infections are hospital-acquired or community-acquired. Therefore, there is a need for effective guidelines to reduce carbapenem resistance in A. baumannii clinical isolates in South Korea.

The widely disseminated CRAB isolates were mostly associated with the CC92 or GC2 genotypes harboring the bla_OXA-23 gene, which is the most common acquired carbapenemase, indicating that the CC92 CRAB has been identified as the major genotypes related to carbapenem resistance worldwide, as well as in Asia.^9,19,20 In South Korea, CC92 CRAB is also responsible for the high resistance rate to carbapenems. According to a 2008 report,²¹ 73.8% (n=299/405) of CRAB clinical isolates were CC92 CRAB. Similar to that in 2017,⁶ the percentage of CC92 CRAB isolates mostly associated with bla_OXA-23 from tertiary hospitals and university hospitals in South Korea was confirmed to increase to 82.0% (n=291/355). In 2023, CC92 CRAB harboring bla_OXA-23 was also the most common genotypes (90.1%, n=519) among the 576 CRAB isolates collected from community hospitals [Table 2]. In addition, according to an observational study that took place during 2001–2020,²² the predominant CRAB genotypes causing invasive infections in children changed from non-CC92 to CC92 as a result of the 2010–2011 outbreaks. According to several of the abovementioned studies, it is considered that the CC92 lineage of CRAB clinical isolates has become the major genotypes throughout South Korea.

In addition to the CC92 CRAB isolates identified in this study, a large number of ST229 CRAB strains belonging to the CC110 (n=47) were isolated from a general hospital in the Seoul region. The ST229 CRAB isolate was originally known as the epidemic clone in 2015 in the Daegu region of South Korea.^23,24 These strains mostly carried ISAbaI-bla_OXA-51, but all the ST229 CRAB isolates in this study carried bla_OXA-23, similar to the findings in 2017.⁶ A total of 47 ST229 multidrug-resistant A. baumannii isolates exhibited the same antimicrobial susceptibility profiles: co-resistant to SAM, AZT, CIP, FEP, GEN, PIP, TMP/SMX, CTX, CAZ, TIM, and TZP. In this regard, the CRAB ST229 isolate seems to be an outbreak caused by clonal spread in a general hospital. Without this clonal spread of ST229 CARB in this study, the ratio of CC92 CRAB isolates would have been slightly higher.

ST373 CRAB (n=7) was the third most prevalent genotype of strains in this study. ST373 A. baumannii isolates are known as hyper-biofilm forming A. baumannii.²⁵ In a previous study, only one ST373 CRAB isolate was detected in South Korea.⁶ Finally, each one of ST397 belonging to CC108, ST447, and ST620 belonging to CC20 CRAB isolates has seldom been reported in South Korea.^6,9,26 The emergence of these rare genotypes is considered to be a distinct characteristic that originated from community-associated infections.

The correct identification of clinical Acinetobacter isolates is important because A. baumannii and non-baumannii Acinetobacter species generally exhibit different AMR pattern.³ Automatic systems such as VITEK-2 may be suitable for antimicrobial susceptibility testing, but the quick and correct detection of Acinetobacter infections is challenging to date.²⁷ Therefore, we confirmed the identification of Acinetobacter species by analyzing the Acinetobacter rpoB gene sequence as the gold standard for molecular analysis. In general, A. baumannii, A. nosocomialis, and A. pittii in order are the most frequently found causative species of healthcare-associated Acinetobacter infections.^3,8 Among the Acinetobacter species in this study, 19 different species were identified, including A. baumannii (n=659), A. pittii (n=55), and A. nosocomialis (n=19), as well as A. courvalinii, A. dispersus, and A. vivianii. Hemolytic and/or proteolytic A. courvalinii, A. dispersus, and A. vivianii of the genus Acinetobacter were previously classified into three putative species terms as genomic species 14BJ, genomic species 17, and taxon 20, respectively.²⁸ To the best of our knowledge, these three strains have never been found in healthcare-associated infections in South Korea. Our findings cannot demonstrate the epidemiological origin of unique and diverse species, but their existence seems to be a feature of community-associated infections. In other words, unlike healthcare-associated infections, various species may be detected in community-associated infections among Acinetobacter species.

For non-baumannii Acinetobacter species, the resistance rate to carbapenems was 8.9% (n=14). These results are similar to those of the 2017 study in South Korea.⁶ These 14 carbapenem-resistant non-baumannii Acineotobacter species carried the three types of carbapenemase (9 bla_OXA-23, 3 bla_NDM-1, and 2 bla_VIM-2) in A. bereziniae, A. nosocomialis, A. pittii, and A. seifertii [Table 1]. The bla_OXA-23 gene, which is mostly associated with A. baumannii or A. nosocomialis, is normally found on the chromosome rather than on a plasmid.²¹ However, reports of A. bereziniae, A. pittii, and A. seifertii isolates harboring bla_NDM-1 and bla_VIM-2 are rare in South Korea.¹¹ In other Asian countries, Pakistan in 2022 report,²⁹ the bla_VIM-type (n=74/200, 37.0%) and bla_NDM-type (n=28/200, 14.0%) for carbapenemases were detected in Acinetobacter calcoaceticus-baumannii complex isolates. Because these bla_NDM-1 and bla_VIM-2 genes are generally found on a plasmid,³⁰ the possibility of AMR dissemination through horizontal transfer across close countries cannot be ignored because Acinetobacter species originating from community-associated infections could be new carriers.

In this study, the total number and geographical distribution of Acinetobacter clinical isolates in South Korea were sufficient to describe the prevalence of the Acinetobacter species. However, a limitation is that the number of isolates collected from the Daejeon, Jeju, and Ulsan regions was so small that it was difficult to represent the exact AMR pattern.

Conclusion

Our results revealed the distribution of Acinetobacter species and revealed a predominance of CC92 CRAB clinical isolates with widespread production of bla_OXA-23 in community hospitals in South Korea. These results were similar to that observed in tertiary hospitals and university hospitals. In addition, we identified rare STs and newly detected species, suggesting that distinct features of Acinetobacter species originating from community-associated infections. Our findings suggest that there is a need for effective methods to reduce carbapenem resistance in A. baumannii in South Korea.