Characterization of Infections with Vancomycin-Intermediate Staphylococcus aureus (VISA) and Staphylococcus aureus with Reduced Vancomycin Susceptibility in South Korea

The aim of the present study was to describe the characteristics of infections with Staphylococcus aureus with reduced vancomycin susceptibility (SARVS) including vancomycin-intermediate S. aureus (VISA) in South Korea, using data from the national sentinel surveillance system during 2014–2016. A total of 66 patients infected or colonized with SA-RVS were reported using the sentinel surveillance system. Among them, VISA was confirmed in 14 isolates (21.2%) and no vancomycin-resistant S. aureus (VRSA) was detected. Most of patients had any kind of indwelling devices (81.8%, 54/66) and underwent surgical procedures in the previous 6 months (84.8%, 56/66). Patients who admitted to an intensive care unit (ICU) in the previous 3 months were 68.2% (45/66). Furthermore, patients who used vancomycin or had MRSA in the previous 1 month were 54.5% (36/66) and 59.1% (39/66), respectively. Upon review of the medical records, 54.5% (36/66) of patients were classified as having SA-RVSassociated infection and 30-day mortality was 19.4% (7/36). Our findings revealed that there was no VRSA in South Korea. SA-RVS including VISA existed particularly in patients who had indwelling devices, history of surgical procedure, and history of ICU admission.

Laboratory procedures. At the laboratory of the NIH, confirmatory testing was performed for vancomycin MIC using agar dilution, broth microdilution, and E-test for all S. aureus isolates sent to the NIH. All isolates had MIC > 2 µg/mL by at least one of those methods and additional antimicrobial susceptibility testing and molecular typing were performed for them.
For molecular typing, genomic DNA was extracted using a Wizard Genomic DNA Purification Kit (Promega, Madison, WI, USA). According to the manufacturer's protocol for bacterial cells, we added lysostaphin at a final concentration of 30 µg/mL in lysis buffer and incubated it at 37 °C for 1 h. Multilocus sequence typing (MLST) was carried out using polymerase chain reaction (PCR) amplification and sequencing of seven housekeeping genes (arc, aroE, glpF, gmk, pta, tpi, and yqiL) using primer pairs as previously described 10 . The allelic profiles and sequence types (STs) were assigned using the MLST website (http://saureus.mlst.net/). Staphylococcal Cassette Chromosome mec (SCCmec) types were identified with the multiplex PCR method 11 . Strains COL, N315, NCCP13860, and MW2 were included as controls for SCCmec types I, II, III, and IV, respectively. Amplification of the spa repeat region was performed using primers spa-1113f (5′-AAGACGATCCTTCGGTGAGC-3′) and spa-1514r (5′-CAGCAGTAGTGCCGTTTGCTT-3′). PCR products were sequenced and the spa types were determined using the Ridom SpaServer 12,13 . Gene responsible for the vanA or vanB were determined using PCR and DNA sequencing method as described previously 14,15 . Classification of S. aureus according to vancomycin MIC. Following the definition of the Clinical and Laboratory Standards Institute (CLSI), S. aureus isolates with vancomycin MIC 4-8 µg/mL were classified as VISA, and those with MIC ≥ 16 µg/mL were classified as VRSA 16 . The remaining reported S. aureus isolates, which showed vancomycin MIC between 2 and 4 µg/mL, were defined as non-VISA SA-RVS.
Classification of patients with VISA or VRSA according to origin of the pathogen. Patients with VISA or VRSA were categorized as cases of "cross-transmission, " "present on admission, " or "sporadic" according to exposure history for the potential origin of the pathogen. A description of each category follows.
www.nature.com/scientificreports www.nature.com/scientificreports/ As for clinical categories, we defined patients who had compatible infectious signs and symptoms (such as bloodstream infection, urinary tract infection, pneumonia, surgical site infection, and so on) as having "SA-RVS-associated infection" based on the US Centers for Disease Control and Prevention/National Healthcare Safety Network (CDC/NHSN) surveillance definition of healthcare-associated infection 17 ; the remaining cases were classified as "colonization with SA-RVS".
Descriptive study of fatal cases of sA-RVs-associated infection. To identify clinical characteristics of deaths caused by SA-RVS-associated infection, we performed a descriptive study. Demographic data, diagnosis on admission, risk factors for VISA or VRSA acquisition, and treatment regimen for SA-RVS-associated infections were summarized. statistical analysis. All statistical analyses were conducted using SPSS version 21.0 for Windows (IBM Corp., Armonk, NY, USA). Categorical variables were analyzed using the chi-square test or Fisher's extract test, as appropriate. Continuous variables were analyzed by independent t-tests. A two-tailed P-value of <0.05 was considered statistically significant. ethics statement. The study protocol was approved by the Institutional Review Boards of the KCDC (2018-02-04-PE-A). The requirement for written informed consent from patients was waived due to the retrospective nature of the study and its impracticability.
As for clinical outcomes, there were no deaths among patients with VISA; however, 13 (43.3%, 13/30) patients with non-VISA SA-RVS died during admission; this difference was not statistically significant (P = 0.680).
Deaths attributed to sA-RVs-associated infection. We analyzed patients who died within 30 days of hospital admission owing to SA-RVS-associated infections ( Table 2). As shown in the table, all deaths occurred among patients with non-VISA SA-RVS.

Discussion
Herein, we report the characteristics of patients carrying SA-RVS including VISA in South Korea. Notably, no VRSA has been isolated in South Korea. The main mechanism for the occurrence of VRSA is via the presence of the vanA gene, which could be transferred from vancomycin-resistant enterococci to S. aureus via plasmids 15,18 . We found that no S. aureus isolates identified via the National Surveillance System for VISA had the vanA or vanB gene.
The presence of a thickened cell wall with an increased number of peptidoglycan layers is mostly applicable mechanism of reduced susceptibility to vancomycin among VISA 19 . The gene mutation that changes the pattern of the cell wall and/or reduces the expression of penicillin-binding proteins in S. aureus is triggered by long-term vancomycin usage 20,21 . In fact, the majority of VISA strains have emerged in patients with MRSA undergoing prolonged vancomycin therapy 19,20 .
Identified risk factors associated with VISA in previous studies include prolonged vancomycin use, previous MRSA colonization, hemodialysis dependence, and use of indwelling devices 19,22 . Furthermore, an observational study in the US revealed that prior vancomycin exposure within 30 days and residence in an ICU were predictors of SA-RVS presence in patients with S. aureus bacteremia 23 . Concordant with previous studies, we found that the majority of patients with SA-RVS used an indwelling device during hospitalization, had history of surgical procedures within 6 months, and admitted to ICU within 3 months.
There was no significant difference in most parameters including 30-day mortality between patients with VISA and patients with non-VISA SA-RVS. A previous study in the US suggested that patients with non-VISA SA-RVS infection had similar clinical characteristics compared to those with VISA infection 22 . Although elevated vancomycin MIC levels in patients with S. aureus infection may cause antibiotic therapy less effective, the association between vancomycin MIC and mortality is still controversial. Some previous meta-analyses have addressed association between vancomycin MIC and worse outcome 24,25 . On the contrary, no statistically significant difference in the mortality between patients with SA-RVS and those with non-SA-RVS were observed in the recent studies 26,27 . A prospective cohort study proposed that reduced vancomycin susceptibility might be linked to reduced disease severity in S. aureus infection 27 . Further studies are necessary to clarify this issue.  www.nature.com/scientificreports www.nature.com/scientificreports/ As for antimicrobial susceptibility, the SA-RVS strains appeared to have retained high susceptibility to linezolid, tigecycline, nitrofurantoin, and trimethoprim/sulfamethoxazole. Fortunately, these agents are available in South Korea and could be treatment options for SA-RVS-associated infections. When comparing VISA isolates with non-VISA SA-RVS isolates, the VISA strains were significantly less susceptible to teicoplanin. A previous report revealed that VISA is well correlated with lower response to teicoplanin because teicoplanin is a glycopeptide and has mechanisms similar to those of vancomycin 28 . Interestingly, two VISA isolates were resistant to daptomycin, which is not available in South Korea. Previous studies have revealed that increased vancomycin MIC in VISA isolates is associated with daptomycin resistance 29,30 . Sakoulas et al. suggested that exposure of S. aureus to vancomycin may affect daptomycin resistance 31 . In our study, two daptomycin-resistant VISA isolates had vancomycin MIC of 4 µg/mL. Another interesting finding was that two VISA isolates were susceptible to oxacillin whereas no non-VISA isolates were susceptible to this agent; both of these isolates carried SCCmec type II. Previous in vitro studies showed that decreased vancomycin susceptibility is associated with increased sensitivity to beta-lactams 32 . Based on such findings, the combination of vancomycin and beta-lactams has been suggested for treatment of SA-RVS infection 33 .

SA-RVS-associated infection
A recent retrospective study demonstrated that the molecular epidemiology of S. aureus is changing in Korea: a community-genotype MRSA strain, ST72, has emerged as a nosocomial pathogen 34   www.nature.com/scientificreports www.nature.com/scientificreports/ and ST5 strains are correlated with hospital-associated MRSA whereas SCCmec type IV and ST72 are correlated with community-acquired MRSA 35,36 . Despite the changing molecular epidemiology, SCCmec type II and ST5 remain predominant among SA-RVS strains in South Korea. Considering that most of our patients had risk factors for acquisition of SA-RVS, most SA-RVS strains might originate in health-care settings in South Korea.
Our study had some limitations. First, we could not acquire sufficient information about antibiotic usage. Basically, the in-depth surveillance in this study was mainly focused on the possibility of cross-transmission within medical institutions, and patients' medical records from previous facilities were not fully available. Thus, we could not draw any conclusions about the relationship between the amount of vancomycin consumption and elevated vancomycin MIC in S. aureus. Second, most of the sentinel medical institutions are tertiary care hospitals and public hospitals. Therefore, the results of the present study are not generalizable to small-or medium-sized hospitals, including long-term care hospitals. Third, confirmatory testing for vancomycin MIC was performed by a variety of methods. E-test tends to yield higher numeric values for MIC than either broth or agar dilution and some isolates showed vancomycin MIC > 2 µg/mL only by E-test. It may have influenced data. Finally, the control group with fully vancomycin susceptible S. aureus was absent, which limits the ability to draw conclusions regarding clinical risk factors for VISA or SA-RVS acquisition.
Despite these limitations, the overall data of this study likely represent a reasonable approximation of the true values, and our findings may represent well the national status of SA-RVS in South Korea.

Conclusion
Our findings revealed that there was no VRSA in South Korea. However, SA-RVS including VISA existed particularly in patients who had indwelling devices, history of surgical procedure, and history of ICU admission. Because only few therapeutic options exist and showing high mortality rate for infections caused by such pathogens, proper preventive strategies for further spread are indispensable. Implementing strict infection-control strategies and antimicrobial stewardship practices in each hospital is required in South Korea.  Table 4. Comparison of antimicrobial susceptibility of S. aureus with reduced vancomycin susceptibility (SA-RVS) isolates by SCCmec types.