Validation of PCR - Based Testing for Surveillance of Drug Resistant Acinetobacterbaumannii ( ISAba1 / bla OXA -51- Like ) in a Pediatric Ward

Objective : This study was performed to validate PCR-based testing for surveillance of Acinetobacter baumannii ( A. baumannii ) in a pediatric ward in response to an outbreak of drug resistant A. baumannii (IS Aba1 / bla OXA-51-like) in our tertiary care hospital. Materials : All children who were admitted to the pediatric ward were included in this study. The surveillance study was carried out from March 2012 to December 2013. Methods : After collecting samples from all children in our pediatric ward, PCR analyses were conducted. IS Aba1 / bla OXA-51-like gene-positive patients were isolated from other patients and strict standard precautionary procedures for infection control were put into practice. Results : During the survey period, a total of 1,038 children underwent surveillance on admission. Of the 16 cases of Acinetobacter baumannii colonization identified during surveillance, 6 were positive for the IS Aba1 / bla OXA-51-like gene, and all IS Aba1 / bla OXA-51-like gene-positive patients received mechanical ventilation. No new IS Aba1 / bla OXA-51-like gene-positive cases were identified during the last six months of surveillance. However, among the 10 IS Aba1 / bla OXA-51-like gene-negative A. baumannii cases, 3 were neonates transferred from the same gynecological hospital. Conclusion : PCR-based testing was effective for stopping transmission of an A. baumannii outbreak in a pediatric ward. This event emphasized the importance of surveillance and interfacility communication among hospitals for prevention of infection outbreaks.


Introduction
Drug-resistant Acinetobacter baumannii (DRA) is an emerging public health threat globally 1) and in Japan 2) . Acinetobacter baumannii (A. baumannii) infections are clinically important because of an association with high in-hospital mortality among elderly people or immunocompromised patients 1) 3) . A. baumannii cannot be detected using ordinary culture methods due to the general classification of Acinetobacter spp. according to genetic type 4) . Once an outbreak occurs, it takes a long time to eradicate A. baumannii due to the persistence of the bacteria in contaminated environments and medical equipment as well as in healthcare personnel with skin colonization 5) 6) . Moreover, treatment options against DRA infections are limited 1) . Therefore, prompt detection and strict adherence to a variety of infection-control measures are essential for containment of an outbreak of A. baumannii 6) .
After detecting an outbreak of A. baumannii among adult patients in our tertiary care hospital in 2012, we began PCR-based testing for Acinetobacter spp. and discovered that the bla OXA-51-like genetic types with gene sequences associated with carbapenem resistance (ISAba1 gene) were responsible for the outbreak. All strains with the same genotype in patients with nosocomial transmission were confirmed to be drug-resistant strains containing this insertion of the ISAba1 gene sequence. In response to this outbreak, active surveillance using PCR was performed in all inpatients of the entire hospital. In our hospital, we treat children in a critical condition at the Emergency and Critical Care Center that was recognized as the origin of the outbreak. There was also a strong possibility of nosocomial transmission of DRA from adult to pediatric patients. Under the aforementioned conditions, we began PCR-based testing for surveillance of DRA in our pediatric ward. The objective of this study was to investigate the effectiveness of surveillance in pediatricsrelated wards during this period and the status of A. baumannii and drug resistance in these wards, with a focus on the prevalence of ISAba1/ bla OXA-51-like gene-positive strains inducing outbreaks in the hospital.

Material and Methods
The following study was performed in a 653-bed tertiary care hospital. The hospitalʼs pediatric ward contains 42 beds, including 3 in the High Care Unit, 6 in the Neonatal Intensive Care Unit (NICU) and 12 in the Growing Care Unit (GCU); the ward has about 1,200 admissions annually. Critically ill children are also admitted to the Pediatric Intensive Care Unit (PICU) in the Emergency and Critical Care Center. Annually about 400 children undergo surgery in the hospital.
ISAba1/ bla OXA51-like gene-positive patients were defined as children admitted to our pediatric ward (including the High Care Unit, Neonatal ICU, Growth Care Unit and PICU) from whom A. baumannii strains with bla OXA-51-like genes with an inserted sequence of ISAba1 had been isolated by means of clinical cultures, over a period of 15 months from March 2012 to May 2013. Data were collected from patient records as well as hospital databases. For all patients, the following were recorded: age; sex; transfer from another institution; the day of isolation of A. baumannii; length of hospital stay prior to isolation of A. baumannii; use of invasive devices; surgery; mechanical ventilation; antimicrobial susceptibility; and antimicrobial drug therapy.
Firstly, we screened Acinetobacter spp. by culture of specimens including the nasal vestibule, pharynx, sputum of transtracheal aspiration from patients with tracheal intubation, injured skin (burns, invasive device insertion site, and wounds, such as surgical wound), urine, and feces. Specimens were collected on admission and weekly thereafter from the same patients. Specimens were submitted to the microbiology laboratory in our hospital for selective cultures on CHROMagar TM Acinetobacter. Samples were cultured on the medium at 35℃ for 18 h, and when red smooth colonies with a diameter of about 2 mm were formed, the bacterium was judged to be Acinetobacter spp.
Patients who were positive for Acinetobacter spp. were isolated from other patients as soon as possible after detection, and strict standard precautionary procedures for infection control were put in place. These included stringent hygiene controls, wearing of disposable isolation gowns, and aggressive cleaning of the environment.
Subsequently, we performed PCR tests on the specimens directly to identify the existence of bla OXA-51-like carbapenemase gene types, which could be used as confirmation of A. baumannii because all A. baumannii strains have intrinsically, and the ISAba1 gene, which could be a positive marker for DRA 5) . An ISAba1-positive strain was defined as an A. baumannii strain with an inserted ISAba1 gene sequence identified by PCR. Detection of the OXA-carbapenemase gene ( bla OXA-51-like) was carried out using multiplex PCR and amplification of ISAba1/ bla OXA-51-like genes was performed using ISAba1 primers and OXA-51-like reverse primers 5) .
Waiting for the culture results, ISAba1 gene-positive patients were isolated in a specially designated ward for ISAba1 gene-positive patients only, whereas ISAba1 gene-negative patients were transferred to single rooms in each ward. Hospital outbreak strain-positive (ISAba1 gene-positive) patients who were readmitted within the period were quarantined in a dedicated ward at the time of admission.
Pulse-field gel electrophoresis (PFGE) was performed for genetic typing of the organisms to confirm consistency of the ISAba1-gene positive strain with the hospital outbreak strain. The sample was treated with a restriction enzyme, Sma I (TAKARA), at 30℃ overnight and subjected to electrophoresis using a CHEF-DRIII system (BIO-RAD). A phylogenetic tree was prepared using Fingerprinting II Software ver. 3.0, and the Tenover criteria were used for PFGE band analysis 7) . Minimum inhibitory concentrations (MICs) were determined by agar dilution following Clinical and Laboratory Standards Institute criteria (using the M100-S25 version) for susceptibility testing. This study was performed by retrospectively extracting data collected in an emergency surveillance of the entire hospital performed in response to the outbreak in our hospital after examination by the infection control committee and approval by the director of the hospital. The Institutional Review Board at Juntendo University Urayasu Hospital approved the study protocol and granted a waiver of informed consent due to the retrospective nature of the study. The medical records of subjects were de-identified prior to analysis.

Results
During the survey period, a total of 1,038 children (excluding overlapping admission) underwent surveillance on admission and 18 cases of Acinetobacter species were identified by selective culture. PCR tests identified these cases as 16 A. baumannii ( bla OXA-51-like gene-positive) strains and 2 nonbaumannii strains. Of the 16 A. baumannii cases, 6 (37.5%) tested positive for ISAba1 gene (ISAba1/ bla OXA-51-like gene-positive), including 1 female and 5 male patients. Two were adolescents (15 and 16 years old) and 4 were infants (0 to 3 years old).
The clinical features of these 6 ISAba1-positive cases are shown in Table- (Figure-1).
A comparison of MICs among A. baumannii strains showed that ISAba1-positive strains had strong resistance to amikacin and levofloxacin, and reduced susceptibility to imipenem (MICs 2-4 μgml -1 ) compared to ISAba1-negative strains. Genotyping by PFGE confirmed that all ISAba1positive cases belonged to the epidemic clone in our hospital. Among ISAba1-negative cases, 3 had the same genotype and were neonates transferred from the same gynecological hospital. The other control cases had independent genotypes (Figure-2).
Complications associated with A. baumannii infection, such as blood stream infection or ventilated associated pneumonia, were not observed.

Discussion
No new ISAba1-positive strain-positive patient was noted in the latter half of the study period, suggesting that the active surveillance was effective in the pediatrics-related wards. Due to the complex epidemiology of A. baumannii 8) , a comprehensive intensified infection control strategy such as effective environmental cleaning 9)-11) , effective sterilization of reusable medical equipment, attention to proper hand hygiene protocols 12)- 14) , and use of contact precautions for infected or colonized patients 15) 16) are required for the containment of a DRA outbreak. Such measures were taken in this study and no new case of ISAba1-positive A. baumannii has been detected in the pediatric ward in the last 6 months of surveillance. The pediatric ward is separated from other wards and entry from other wards is limited. This made it easier to take measures against infection, and convergence based on active surveillance were achieved earlier. As a result, we have had no newly colonized patients apart from those transferred from other hospitals. However, some ISAba1-negative cases appeared persistently despite the fact that  additional measures for infection control were also taken. That would suggest that such measures were not sufficient, and similar measures against ISAba1-negative A. baumannii may be required in such cases. In addition, since colonization may be brought into a hospital from outside, surveillance may be necessary not only for a given hospital but also for hospitals in surrounding areas.
In the current study, we found that the ISAba1positive strains transmitted in our pediatric ward were highly resistant to aminoglycosides and new quinolones, but had reduced susceptibility to imipenems (MICs 2 μgml -1 ). On the other hand, the incidence of resistance was relatively low in ISAba1-negative cases. Of the β-lactamases in A. baumannii, those with carbapenemase activity are of most concern. The OXA type carbapenemase is a characteristic of a phenotype that is difficult to discriminate clearly, and confirmation using PCR is necessary 4) . The bla OXA-51-like gene occurs naturally in A. baumannii strains, but its role in carbapenem resistance appears to be related to the presence of the ISAba1 gene. ISAba1 contains promoter regions that lead to overexpression of downstream resistance determinants 17) , and ISAba1 also seems to be important in expression of resistance to other antibiotics and is one of the principal causes of DRA 4) . The ISAba1-positive strain in this study was an identical clone to the hospital outbreak strain. The strain showed resistance to gentamycin and levofloxacin, but not to imipenem, although sensitivity was decreased. This strain may be strongly resistant and may become multi-drug resistant, for which continuous strict surveillance is necessary.
The High Care Unit was found to be a significantly contaminated site, but only one case was isolated from the Emergency and Critical Care Center, which was the origin of the ISAba1-positive strain in adult patients. This suggests a strong possibility of transportation of the ISAba1-positive strain from the Emergency and Critical Care Center to the High Care Unit via medical staff. Horizontal infection from ISAba1-positive cases could be prevented in pediatric wards, but the risk of horizontal infection from adult carriers should be considered in an Emergency and Critical Care Center, and the status of bacteria should be investigated in both pediatric and adult patients. Support with mechanical ventilation seems to be a major risk factor for infection by A. baumannii in adults 6) 18) and most ISAba1 gene-positive pediatric patients in our study had undergone tracheotectomy and were on a ventilator. In contrast, tracheal intubation was applied in only 40% of ISAba1-negative cases. A previous study found that pediatric patients on mechanical ventilation in a surgical intensive care unit were at risk for carbapenemresistant A. baumannii 19) . Although we did not identify any multidrug resistant A. baumannii cases in the current study, the findings indicate that mechanical ventilation is also likely to be a risk for DRA, such as in our cases (ISAba1/ bla OXA-51), among all hospitalized children. Further studies are needed to confirm the risk factors in children.
In conclusion, PCR-based testing for A. baumannii targeting the ISAba1/ bla OXA-51-like gene was effective for controlling an outbreak in our particular pediatric ward. However, while we were able to stop transmission of the ISAba1-positive strain, which was resistant to amikacin and levofloxacin, by introducing active surveillance in our hospital, one crucial point to note is the possibility of transportation of A. baumannii among hospitals in the same region. This highlights the importance of ongoing surveillance and interfacility communication.