Aeromonas dhakensis: Clinical Isolates with High Carbapenem Resistance

Aeromonas dhakensis is ubiquitous in aquatic habitats and can cause life-threatening septicaemia in humans. However, limited data are available on their antimicrobial susceptibility testing (AST) profiles. Hence, we aimed to examine their AST patterns using clinical (n = 94) and non-clinical (n = 23) isolates with dehydrated MicroScan microdilution. Carbapenem resistant isolates were further screened for genes related to carbapenem resistance using molecular assay. The isolates exhibited resistance to imipenem (76.9%), doripenem (62.4%), meropenem (41.9%), trimethoprim/sulfamethoxazole (11.1%), cefotaxime (8.5%), ceftazidime (6%), cefepime (1.7%) and aztreonam (0.9%), whereas all isolates were susceptible to amikacin. Clinical isolates showed significant association with resistance to doripenem, imipenem and meropenem compared to non-clinical isolates. These blacphA were detected in clinical isolates with resistance phenotypes: doripenem (67.2%, 45/67), imipenem (65.9%, 54/82) and meropenem (65.2%, 30/46). Our findings showed that the MicroScan microdilution method is suitable for the detection of carbapenem resistance in both clinical (48.9–87.2%) and non-clinical (4.3–13.0%) isolates. This study revealed that A. dhakensis isolates had relatively high carbapenem resistance, which may lead to potential treatment failure. Continued monitoring of aquatic sources with a larger sample size should be carried out to provide further insights.


Introduction
Aeromonas species are ubiquitous, Gram-negative, facultative anaerobes, which can cause a variety of infections in poikilothermic animals and humans [1]. They can be isolated from virtually all environmental niches where bacterial ecosystems exist (aquatic habitats, fish, foods, domesticated pets, invertebrate species, birds, ticks, insects and natural soils). The number of species in the genus increased rapidly in the age of molecular genetics and consists of 36 species reported to date [2]. The exact incidence of Aeromonas infection in humans on a global basis is limited, as many cases are undetected or not reported.
Aeromonas dhakensis (previously named A. hydrophilla or A. aquariorum) is emerging as a clinically important pathogen that can cause severe soft tissue degloving infections arising from occupational and recreational hazards, as well as bloodstream infections in immunocompromised individuals with malignancy and cirrhosis [3]. A higher mortality rate was observed in A. dhakensis bacteraemia compared to bacteraemia caused by non-A. dhakensis species [4]. Researchers reported greater virulence properties in A. dhakensis than other species, such as more robust biofilm formation and a lower survival rate in Caenorhabditis elegans, and some strains exhibited a higher minimum inhibitory concentration (MIC) for certain antimicrobial agents [4]. Recently, two fatal cases of A. dhakensis bacteraemia and necrotising fasciitis in severe dengue patients were reported in Southern Taiwan, and a fatal case of A. dhakensis septicaemia in a hepatitis B virus-infected patient after the ingestion of a meal of raw snakehead fish was reported in China [5,6].
A. dhakensis has been misidentified as A. hydrophila or A. caviae by phenotypic methods in the past decade. Our previous studies conducted since 2012 identified A. dhakensis as the predominant species (47/94, 50%) among clinical isolates in Malaysia by using a combination gene analysis of GCAT and rpoD genes, as well as multilocus sequence typing [7,8]. Following these studies, the presence of A. dhakensis was reported from aquatic sources in Malaysia, including multipurpose freshwater recreational lakes in Selangor [9], tank water of ornamental fish in Klang Valley [10], food fish in East Malaysia [11] and as clinical isolates of A. dhakensis in Singapore [12]. All aforementioned studies shed light on the clinical relevance of A. dhakensis and its ability to present/colonise various sources posing public health concerns. Recognising its clinical relevance and capability of causing invasive disease, understanding the antimicrobial resistance (AMR) profile of A. dhakensis is important to support the selection of optimal treatment regimens. In this study, we examined the antimicrobial resistance patterns of A. dhakensis from clinical and non-clinical sources (ornamental fish tank water, freshwater from a recreational lake and food fish) based on minimum inhibitory concentrations (MICs) using the broth microdilution method.

Discussion
A. dhakensis, increasingly recognised as an emerging human pathogen, was identified as the predominant Aeromonas species among clinical isolates in Malaysia (50%, 47/94) and Singapore (40.5%, 47/116). The similarity of dominant species in the two countries may be due to comparable environmental factors, such as climate, temperature and humidity, as well as dietary patterns. In this study, although the 94 clinical isolates recovered from patients in Malaysia and Singapore were previously characterised through molecular fingerprinting as not clonally related, A. dhakensis isolates from both countries were found to harbour a higher number of virulence genes when compared to other Aeromonas species [12]. All clinical isolates were positive for at least 6 of the 15 virulence genes, and these genes (ela, lip, alt, ser, exu, fla and aer) were present in ≥90% of the isolates (Supplementary Figure S1). The dominance of A. dhakensis infections in Malaysia and Singapore was postulated to be an indication of the virulence potential of this species, in accordance with findings of other studies [13,14]. Furthermore, A. dhakensis in both countries were revealed to be not unequivocally considered diarrhoeal pathogens, as they were frequently isolated from extraintestinal sites, i.e., most frequently from pus/wound culture in Malaysia clinical isolates and from blood culture from Singapore clinical isolates [12]. Whereas antimicrobials are usually not prescribed for self-limiting diarrhoeal infections, empirical clinical antimicrobial therapy may be considered for severe diarrhoea (e.g., bloody diarrhoea) or invasive infections (e.g., septicaemia). Recognising the clinical relevance of A. dhakensis and its ability to cause invasive disease, the antimicrobial resistance profiles of the A. dhakensis from clinical and non-clinical sources generated from this study could facilitate a better understanding of the antimicrobial resistance traits of the pathogen to support the selection of optimal treatment regimens for disease management.
Carbapenems are recognised as a critically important antimicrobial (CIA) by the World Health Organization for human medicine. Resistance to such antimicrobials is a public health concern, as it would render treatment of infection less effective and increase healthcare costs. The mortality rate of patients infected with carbapenem resistant Aeromonas spp. due to meropenem treatment failure was reported as 33.3% (7/21) and 100% in patients with bacteraemia [15]. Further genotypic screening revealed that more than 65% of carbapenem resistant clinical isolates carried the chromosomally encoded metallo-beta-lactamase bla cphA , suggesting an intrinsic AMR mechanism could underlie the local epidemiology of resistant Aeromonas infections in the investigated countries. However, the clinical isolates included in our study were from selected public hospitals and might not represent complete case data to draw a concrete conclusion on resistance patterns from both countries. In the future, more samples from more hospitals in these countries could be screened to further substantiate our current findings.
Antimicrobial resistance of clinical A. dhakensis to imipenem and meropenem has been reported previously in two studies in Taiwan but with lower frequencies compared to this study [4,13]. Chen et al. reported that the resistance rate of imipenem was 10.8% (4/37) using E-test strips [4]. Wu et al. documented a resistance rate of 4.2% (2/48) for both imipenem and meropenem using the broth microdilution method of the Trek Sensititre system [13]. The relatively higher carbapenem resistance rate in the clinical isolates in our study has three possible explanations:(i) different platforms with different inoculum preparations; (ii) geographically related variations in antimicrobial resistance patterns due to differences in the healthcare system, particularly in terms of treatment regimen recommended, practices of healthcare professionals and patients' behaviour towards the use of antimicrobials; or (iii) indirect factors, such as lifestyle, seasonal difference, dietary preference, e.g., the Southeast Asia region exhibits a remarkably high per capita fish consumption of 33.4 kg compared to Asia as a whole (21.3 kg) [16]. This calls for constant monitoring of A. dhakensis from aquatic-related sources for public health mitigation. It also underscores that clinical empirical carbapenem therapy for A. dhakensis infection should be used with caution.
Among non-clinical isolates, those recovered from the gills of healthy food fish Marble Goby exhibited a relatively higher resistance rate to carbapenems than that of others (ornamental fish tank water and recreational lake water): imipenem, 100% vs. 16.7% and 100%; doripenem, 75% vs. 16.7% and 0%; and meropenem, 25% vs. 11.1% and 0% (Data not shown). The occurrence of A. dhakensis (12.7%, 8/63) was also reported in freshly and moribund Nile tilapia samples in India [17]. The study reported that one A. dhakensis isolate randomly selected for AMR profiling showed resistance towards imipenem and meropenem with an MIC >32 µg/mL via MIC strips (HiMedia), whereas, in our study, we determined the MIC of the carbapenems (imipenem and meropenem) to be up to 8 µg/mL. The detection of carbapenem resistant A. dhakensis in food fish revealed a potential risk to public health, as well as a potential pathway for humans to acquire resistant bacteria through consumption of aquaculture food contaminated with resistant bacteria. However, only a limited number of isolates (n = 4) from a relatively small sample size of food fish (n = 15) collected at intervals from November 2015 to October 2017 were used in the study. Further longitudinal monitoring for AMR A. dhakensis with larger sample size and increased sampling coverage, preferably complemented by whole-genome sequencing, should be carried out to identify possible transmission routes of A. dhakensis. The exceptional rate of aquaculture growth in Asia as a region should also be considered in terms of meeting the majority of the global seafood demand. Intensification of the aquaculture system might be expected to drive the usage of antimicrobials to maintain animal health. Hence, continued monitoring of resistant aeromonads is warranted using a larger sample size in the environmental niches that the bacterium possibly inhabits.
Among the three carbapenems, the AST for meropenem of the 94 clinical isolates and 4 non-clinical isolates from food fish (Marble Goby) was previously reported using the Kirby-Bauer disk diffusion method [11,12]. However, in these studies, only one meropenem resistant clinical isolate was detected, and no resistance was observed in food fish isolates. Possible explanations for the discrepancy between the results of microdilution and disk diffusion assays could be the concentrations of antimicrobial agents and different inoculum preparation methods. MicroScan Prompt (the microdilution method used in this study) preparation offers a standardised inoculum (generally expected inoculum size: 6.9 × 10 5 cfu/mL), whereas the disk diffusion test requires manual preparation with inoculums adjusted to match a 0.5 MacFarland turbidity standard (10 4 -10 5 cfu/mL). The determination of in vitro susceptibility of carbapenemase-producing aeromonads by disk diffusion or dilution techniques usually yields a susceptible genotype unless using a large inoculum (3 × 10 8 cfu/mL) [18]. For the detection of meropenem resistance in A. dhakensis, Sinclair et al. reported a low accuracy of disk diffusion (2.6%, 1/39) and E-test (0%) compared to the microdilution method (61.5%, 24/39) [19]. By using the microdilution method via the MicroScan system, the emergence of carbapenem resistance of Aeromonas spp., including A. dhakensis, was reported in Colombia [15]. This observation underscores the importance of selecting an appropriate method for the examination of carbapenem resistance in Aeromonas.
Each AST detection platform has inherent strengths and limitations. As shown in this study, most MIC results from MicroScan for nine antimicrobial agents-piperacillin/ tazobactam, cefuroxime, cefoxitin, ertapenem, gentamicin, tetracycline, ciprofloxacin, levofloxacin and chloramphenicol-were outside the reference MIC range of CLSI. These MIC results were categorised as "not determinable", as the MICs could not be assessed using CLSI 2015 breakpoints (Table 3). A high rate of "not determinable" was observed in cefoxitin (82.9%) and ertapenem (62.4%) ( Table 1). This also calls for continuous improvement of commercial platforms to incorporate the latest antimicrobial breakpoints [20].
The kappa statistics for A. dhakensis showed almost perfect agreement (kappa value: 0.810-1.000) between CLSI 2015 and EUCAST 2020 for cefepime, ceftazidime and trimethoprim/sulfamethoxazole. There is no interpretative breakpoint for carbapenem resistance provided by EUCAST 2020; thus, the agreement of carbapenem resistance between EUCAST and CLSI methods was not discussed in this study. For ciprofloxacin and levofloxacin, kappa analysis revealed no agreement (kappa value = 0), as four clinical isolates with MIC >2 µg/mL interpreted as susceptible to ciprofloxacin by CLSI 2015 were categorised as resistant by EUCAST 2020. On the other hand, a clinical isolate with >4 µg/mL interpreted as susceptible to levofloxacin by CLSI 2015 was categorised as resistance by EUCAST 2020. Overall, EUCAST 2020 has a more stringent breakpoint for susceptibility for selected antimicrobial agents compared to CLSI 2015 ( Table 3). The stringent breakpoint might help to curb the inappropriate use of antibiotics and control the rising rate of AMR, but this may require further harmonisation, as this has implications for laboratories considering switching between CLSI and EUCAST, as well as for a large-scale AMR surveillance comparing data within and between countries.
Further investigation of the genetic background of carbapenem resistant isolates in our study revealed a chromosomally encoded metallo-β-lactamase cphA gene, indicating that these clinical isolates are attributed to intrinsic resistance. This finding is in agreement with previously published reports on A. dhakensis isolated from human infections in Australia [19], Taiwan [20] and China [21]. The results suggest that clinical use of carbapenem monotherapy should be considered with caution in order to avoid potential treatment failure and that alternative antimicrobial treatment options, such as fourth-generation cephalosporins, quinolones, amikacin, aztreonam and trimethoprim/sulfamethoxazole, remain active for most A. dhakensis isolates.

Bacterial Isolates
A total of 117 A. dhakensis isolates were retrieved from previous studies [7][8][9][10][11][12]22]. These isolates consisted of 94 clinical isolates from Malaysia and Singapore and 23 non-clinical isolates from Malaysia (Table 4). Briefly, these isolates were recovered using Aeromonas selective agar, and their genus identities were confirmed using GCAT gene and species level using rpoD gene sequencing. For non-clinical strains, non-replicate isolate clones were confirmed using enterobacterial repetitive intergenic consensus PCR. The bacterial isolates were revived from glycerol stock onto 3 mL of LB broth and incubated at 35 • C for 20 h with agitation. The next day, bacterial suspensions were subcultured on LB agar and incubated at 35 • C for 20 h.

Antimicrobial Susceptibility Testing
The isolates were subjected to antimicrobial susceptibility testing by the broth microdilution using a MicroScan plate (Beckman Coulter, CA, USA). Four well-isolated colonies were selected using an inoculation wand of the Prompt Inoculation System D and inoculated into a diluent. Following mixing, 115 µL of bacterial suspension was added to each well of the MicroScan plate. The plates were incubated at 35 • C for 20 h, and MIC values were determined manually, followed by susceptibility categorisation based on the CLSI and EUCAST guidelines [23,24]. Escherichia coli ATCC 25922 and Pseudomonas aeruginosa ATCC 27853 were used as quality-control organisms as per the manufacturer's protocol.

Screening for Carbapenem-Resistant Genes
The genomic DNA of carbapenem resistant A. dhakensis isolates was extracted using the boiling method [25]. These isolates were screened for seven carbapenem-related genes, including bla cphA , bla KPC , bla VIM , bla NDM , bla GES-24 , bla IMP-19 and bla OXA-48 , using primers and conditions as described in published studies (Supplementary Table S1). A representative of a positive isolate was randomly chosen and confirmed by direct DNA sequencing.

Conclusions
In summary, our study revealed a high prevalence of carbapenem resistance in A. dhakensis clinical isolates, and most of these isolates were found to harbour the bla cphA gene, which is linked to intrinsic resistance. Given their inherent nature, A. dhakensis infections could be difficult to be treated with carbapenem monotherapy before screening. Therefore, the reliable identification of carbapenemase-producing isolates is an important first step in prescribing an appropriate drug in order to prevent the development of untreatable infections.