A. baumannii-A. calcoaceticus complex includes opportunistic pathogens affecting severely ill patients. The choice of treatment for serious infections caused by Acinetobacter is frequently based on the use of carbapenems1. However, carbapenem resistance among clinical isolates of Acinetobacter spp. has been reported worldwide. In addition, these isolates may have concomitant resistance to most conventional antimicrobial agents and cause difficulty in treating infections, and leave few treatment options2.
Different mechanisms can confer resistance to carbapenems in Acinetobacter, such as decreased permeability of the outer membrane, increased expression of efflux pumps, changes in the affinity of penicillin-binding proteins, and the production of carbapenemases. Among them, the production of carbapenemases, such as more frequent Class D beta-lactamases, also called oxacillinases (OXAs), less frequent Class B, also known as metallo-beta-lactamases (including IMP, VIM, SIM, and NDM-1 types) and Class A (of KPC or GES type) is the most important carbapenem resistance mechanism3–5.
The main groups of OXA-type carbapenemases identified in A. baumannii are OXA-23-like, OXA-24/40-like, OXA-58-like, OXA-143-like, and OXA-235-like groups and are composed of acquired enzymes and OXA-51-like group, which encodes a chromosomal intrinsic OXA in A. baumannii. This intrinsic OXA may be super-expressed and associated with carbapenem resistance6,7. In Brazil, carbapenem-resistant Acinetobacter usually produce OXA-23, followed by OXA-1438,9.
Thus, we aimed to characterize carbapenem-resistant Acinetobacter isolates obtained from two health institutions located in the city of Niterói, Rio de Janeiro State, Brazil, evaluated the production of OXAs, and determined their genetic relationship.
One hundred and fourteen Acinetobacter isolates were obtained from patients at two general hospitals located in Niterói; Hospital 1 (a 290-bed public university teaching hospital; n = 61) and Hospital 2 (a 201-bed tertiary private hospital; n = 53) from March to December 2013.
The most frequent sites for the collection of these specimens were the lower respiratory tract containing secretions (48; 42.1%) such as tracheal aspirate (40; 83.3%), bronchoalveolar lavage (6; 12.5%), sputum and pleural fluid (1 each; 2.1% each), the blood (23; 20.2%), the urine (12; 10.5%), and catheter tip (9; 7.9%). Other sources represented 11.4% (n = 13) and the origin of nine (7.9%) isolates could not be determined.
Identification and antimicrobial susceptibility testing
The isolates were previously identified using VITEK-2 automated system (BioMerieux, Craponne, France) as A. baumannii-A. calcoaceticus complex at the microbiology laboratories of each of the two health institutions. The final identification of A. baumannii-A. calcoaceticus complex was performed using matrix-assisted laser desorption ionization-time of flight mass spectrometry on a Maldi Biotyper platform (Bruker Daltonics, Germany).
The disc diffusion method was used to evaluate susceptibility to the following antimicrobials agents according to Clinical and Laboratory Standards Institute guidelines10: amikacin (30 μg), ampicillin/sulbactam (10/10 μg), ceftazidime (30 μg), cefepime (30 μg), ciprofloxacin (5 μg), gentamicin (10 μg), imipenem (10 μg), meropenem (10 μg), sulfamethoxazole/trimethoprim (23.75 μg/1.25 μg), tetracycline (30 μg), and tobramycin (10 μg). Minimum inhibitory concentration (MIC) values for imipenem were determined by Etest (AB Biodisk, Solna, Sweden).
The isolation of bacterial DNA was accomplished through thermal lysis, according to the methodology proposed by Schuenck with modifications11. PCR was used to detect the genes encoding OXA-like carbapenemases: bla OXA-23-like12, bla OXA-24-like12, bla OXA-58-like12, bla OXA-51-like12, and bla OXA-14313.
The analysis of genetic diversity was investigated through pulsed-field gel electrophoresis (PFGE), as described in a previous study14 using ApaI restriction enzyme (Invitrogen, São Paulo, Brazil). For PGFE analysis, 38 OXA-23 producing isolates from patients in both hospitals (n = 25, Hospital 1 and n = 13, Hospital 2) were included. Only one isolate per patient was considered. Seven representative strains of the two major clonal groups detected in Hospital 1 (genotype A, n = 4 and genotype B, n = 3) in a previous study, with OXA-23-producing Acinetobacter isolates from the year 2007 to 2009 (LL Corrêa, CRV Mendonça-Souza; data unpublished) were also included in the analysis for comparison purposes. The patterns obtained were analyzed using BioNumerics v7.6 software (Applied Maths, Belgium). The comparison of the banding patterns was accomplished through the unweighted pair-group method with arithmetic mean, with a tolerance and optimization of 1.0% using the Dice correlation coefficient. A similarity cut-off of 80% was used for clustering isolates in the same genotype.
The results of antimicrobial susceptibility testing showed high resistance rates of most antimicrobials tested. Among Hospital 1 isolates, the highest resistance rates were observed with imipenem and meropenem (92%), followed by ciprofloxacin (90%), whereas among Hospital 2 isolates, the highest resistance rates were seen with imipenem and ciprofloxacin (100%), followed by meropenem (92%). The highest rates of susceptibility were observed with amikacin (3/114; 2.6%) and tetracycline (4/114; 3.5%) in both hospitals. In general, the resistance rates verified in Hospital 2 were higher than those in Hospital 1, except for trimethoprim/sulfamethoxazole and tobramycin. In addition, given the number of carbapenem-resistant isolates from both hospitals, 82.1% isolates were resistant to at least one agent of two more other classes of antimicrobials tested, indicating multidrug resistance.
According to the PCR results, 96.5% (110/114) of isolates presented bla OXA-51-like, originally intrinsic to A. baumannii, and 95.6% (109/114) carried bla OXA-23-like. No isolate was positive for the other OXA-encoding genes investigated.
All 109 bla OXA-23-like/bla OXA-51-like-positive isolates were resistant to at least one of the carbapenems tested, except for one isolate (from Hospital 1) that had an imipenem MIC of 0.38 μg/mL. One isolate positive to only bla OXA-51-like (from Hospital 2) was resistant to 9 of 11 antibiotics tested including the two carbapenems tested, and was susceptible to only tetracycline and ceftazidime.
PFGE analysis of the 25 isolates from Hospital 1 revealed a polyclonal pattern, but with three major genotypes: A (n = 5; 20%), C (n = 4; 16%), and D (n = 4; 16%). The five isolates belonging to genotype A clustered with four genotype A representative strains, detected in a previous study, with a coefficient of similarity of ≥ 80%. None of the analyzed isolates was related to the three representative strains belonging to genotype B. (Table 1).
For Hospital 2, one predominant genotype, named N, included six isolates (46.2%), clustered with 80% similarity (Table 1). Inter-hospital dissemination was not observed in our study.
|Pulsotype||Isolate||Hospital||Date of Isolation||Clinical specimen||Antimicrobial resistance profilea|
|A||CS30122||1||04/30/13||Blood||IMP; MER; ASB; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30134||1||05/14/13||Urine||IMP; MER; ASB; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30151||1||05/24/13||Liquor||IMP; MER; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30153||1||05/26/13||Liquor||IMP; MER; ASB; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30246||1||09/02/13||Blood||IMP; MER; CPM; SUT|
|C||CS30105||1||03/25/13||Catheter tip||IMP; MER; GEN; CIP; SUT|
|CS30115||1||04/19/13||Catheter tip||IMP; MER; CIP; SUT|
|CS30121||1||05/01/13||Blood||IMP; MER; GEN;TOB; CIP; SUT|
|CS30176||1||06/20/13||Urine||IMP; MER; CPM; CIP; SUT|
|D||CS30279||1||10/01/13||BALb||IMP; MER; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30285||1||10/07/13||Blood||IMP; MER; CPM; GEN; TOB; CIP; SUT|
|CS30289||1||11/01/13||Blood||IMP; MER; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30292||1||11/02/13||Blood||IMP; MER; CIP; SUT|
|E||CS30321||1||11/08/13||Urine||IMP; MER; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30322||1||11/23/13||Blood||IMP; MER; CAZ; CIP; SUT|
|F||CS30118||1||04/19/13||Tracheal secretion||IMP; MER; CAZ; CIP|
|G||CS30278||1||09/30/13||Blood||IMP; MER; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30282||1||09/19/13||Skin biopsy||IMP; MER; CAZ; CPM; GEN; TOB; CIP; SUT|
|H||CS30280||1||09/30/13||Urine||IMP; MER; CIP|
|CS30283||1||09/30/13||Blood||IMP; MER; CIP|
|I||CS30204||1||07/09/13||Blood||IMP; MER; CAZ; CPM; GEN; TOB; CIP; SUT|
|J||CS30177||1||06/21/13||Catheter tip||IMP; MER; CPM; SUT; CIP|
|K||CS30323||1||12/13/13||Urine||IMP; MER; CAZ; CPM; GEN; TOB; CIP; SUT|
|L||CS30284||1||10/14/13||Blood||IMP; MER; CPM; CIP; SUT|
|M||CS30174||1||06/07/13||Renal perfusion fluid||none|
|N||CS30104||2||03/25/13||Catheter tip||IMP; MER; ASB; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30162||2||05/25/13||Blood||IMP; MER; CIP; SUT|
|CS30199||2||07/04/13||Abdominal fragment||IMP; MER; CAZ; CPM; GEN; CIP; SUT|
|CS30195||2||07/29/13||BAL||IMP; MER; CAZ; CPM; GEN; TOB; CIP; SUT|
|CS30213||2||07/30/13||Blood||IMP; MER; ASB; CAZ; CPM; GEN; TOB; CIP; SUT|
|t||CS30214||2||07/30/13||BAL||IMP; MER; ASB; CAZ; CPM; GEN; TOB; CIP; SUT|
|O||CS30111||2||05/20/13||BAL||IMP; MER; ASB; CPM; CIP; SUT|
|P||CS30182||2||06/01/13||Catheter tip||IMP; CAZ; CPM; GEN; CIP|
|Q||CS30197||2||07/15/13||BAL||IMP; MER; CAZ; GEN; TOB; CIP; SUT|
|R||CS30254||2||09/09/13||Tracheal secretion||IMP; MER; ASB; CAZ;CPM; GEN; CIP; SUT|
|S||CS30258||2||09/11/13||Rectal swab||IMP; MER; CIP; SUT|
|T||CS30136||2||04/29/13||Pleural fluid||IMP; MER; CAZ; CPM; CIP|
|U||CS30198||2||07/02/13||Catheter tip||IMP; MER; CIP|
a IMP: Imipenem, MER: Meropenem, ASB: Ampicillin/Sulbactam, CAZ: Ceftazidime, CPM: Cefepime, GEN: Gentamicin, TOB: Tobramycin, CIP: Ciprofloxacin, SUT: Sulfamethoxazole/Trimethoprim; b: BAL: Bronchoalveolar lavage.
This study described the resistance profiles and genetic relatedness of carbapenem-resistant A. baumannii complex isolates collected from patients in two health institutions located in Niterói, Rio de Janeiro, Brazil.
Carbapenem resistance is a serious problem in the treatment of infections caused by Acinetobacter, since these antibiotics are considered as one of the best therapeutic options for the treatment of severe infections caused by Acinetobacter. A study from SENTRY showed an increase in the proportion of carbapenem-resistant Acinetobacter spp., with the rate of 70% in 2008-200915. Another study from Niterói also verified carbapenem-resistance rate of 70% among Acinetobacter isolated in 2007-200916. In the present study, an even higher resistance rate to carbapenems was observed (> 90%) comparable to the results of other recent Brazilian studies that also showed carbapenem resistance rate > 90 % amongA. baumanniiisolates17,18, highlighting an increasing trend in the dissemination of carbapenem-resistant Acinetobacter isolates in the last few years.
We noted that carbapenem resistance was mediated by the enzyme OXA-23 in most isolates. These results are in agreement with the literature, which reported the spread of OXA-23-producing Acinetobacter strains in various locations worldwide and the predominance of this OXA in Brazilian territories, being directly responsible for the high rates of carbapenem resistance8,19–21. The relationship between OXA-23 production and imipenem-resistance ratio among the isolates in both hospitals was also observed. The isolates positive for bla OXA-23 were also resistant to imipenem, except one isolate.
Furthermore, the high prevalence of bla OXA-51 positive isolates detected in this study showed that A. baumannii was the most frequent species, since the oxacillinase is intrinsic in this species6,22. It is noteworthy that, although some studies have described the occurrence of bla OXA-51 in A. nosocomialis, this still appears to be a rare event22.
A carbapenem-resistant isolate was positive only for bla OXA-51. The reduced susceptibility to carbapenems in this non-OXA-23 isolates may be mediated by bla OXA-51 overexpression due to the association with the IS element23. One carbapenem-susceptible isolate was positive for bla OXA-23 and bla OXA-51. This result may be explained by the absence of IS element upstream in the bla OXA-23-like24,25. However, susceptible carbapenem isolates carrying bla OXA-23 are considered as silent reservoirs of this gene and can be the source of their spread in a nosocomial environment26.
In this study, bla OXA genes other than bla OXA-23 and bla OXA-51 were not found, but a bla OXA-72-positive Acinetobacter clinical isolate was obtained for the first time from a public hospital in Niterói, Rio de Janeiro. The increasing occurrence of OXA-72-positive Acinetobacter isolates in Brazil and OXA-143-producing isolates highlights the importance of continuous epidemiological surveillance to help prevent the dissemination of these organisms9,18.
A wide varieties of clonal lineages of bla OXA-23 A. baumannii isolates causing hospital outbreaks has been reported in various studies in Brazil27,28. In this study, genotype A that was detected previously was observed in 5 of the 25 OXA-23-positive strains from Hospital 1 analyzed by PFGE. These results indicate that this strain circulates in the hospital. Some of the OXA-23-producing clones disseminated in Brazil as the clones belonging to ST79 persist and disseminate in the hospital environment18. In this study, we did not perform Multilocus Sequence Type MLST analysis of the OXA-23-positive isolates; therefore, our conclusions about these results obtained are limited.
Among the 13 OXA-23-positive isolates from Hospital 2 that was analyzed by PFGE, a multidrug-resistant predominant genotype (N) was detected, which included 46.2% of the isolates. This result suggests that the intra-hospital spread of this particular clonal group may have contributed to the high rate of carbapenem resistance observed in this institution.
Isolates with unique profiles were detected in both hospitals, indicating that in addition to clonal spread, bla OXA-23 was also acquired through horizontal spread in the investigated population. Studies have shown that the dissemination of carbapenemase-producing isolates appears to be due to clonal dissemination; however, studies have also highlighted the horizontal spread of bla OXA-2329,30.
In conclusion, carbapenem resistance mediated by OXA-23 was high, with most isolates being resistant to carbapenems. A high genetic diversity was verified among the OXA-23-positive isolates analyzed, with the occurrence of both clonal and horizontal dissemination of bla OXA-23-like. These results suggest a need for continuous epidemiological surveillance studies to assist in the control of the dissemination of these carbapenem-resistant strains in the investigated hospitals.