Molecular analysis of methicillin-resistant Staphylococcus aureus dissemination among healthcare professionals and / or HIV patients from a tertiary hospital

Introduction: Methicillin-resistant Staphylococcus aureus (MRSA) is a nosocomial pathogen in community settings. MRSA colonized individuals may contribute to its dissemination; the risk of MRSA infection is increased in human immunodefi ciency virus/acquired immune defi ciency syndrome (HIV/AIDS) patients, although the prevalence of colonization in this group is not well established. The present study addressed this issue by characterizing MRSA isolates from HIV/AIDS patients and their healthcare providers (HCPs) to determine whether transmission occurred between these two populations. Methods: A total of 24 MRSA isolates from HIV-infected patients and fi ve from HCPs were collected between August 2011 and May 2013. Susceptibility to currently available antimicrobials was determined. Epidemiological typing was carried out by pulsed-fi eld gel electrophoresis, multilocus sequence typing, and Staphylococcus cassette chromosome (SCCmec) typing. The presence of heterogeneous vancomycin-intermediate Staphylococcus aureus (hVISA) and heterogeneous daptomycin-resistant Staphylococcus aureus (hDRSA) was confi rmed by population analysis profi le. Isolates characterized in this study were also compared to isolates from 2009 obtained from patients at the same hospital. Results: A variety of lineages were found among patients, including ST5-SCCmecII and ST30-SCCmecIV. Two isolates were Panton-Valentine leukocidin-positive, and hVISA and hDRSA were detected. MRSA isolates from two HCPs were not related to those from HIV/AIDS patients, but clustered with archived MRSA from 2009 with no known relationship to the current study population. Conclusions: ST105-SCCmecII clones that colonized professionals in 2011 and 2012 were already circulating among patients in 2009, but there is no evidence that these clones spread to or between HIV/AIDS patients up to the 7th day of their hospitalization.

Methicillin-resistant Staphylococcus aureus (MRSA) is a leading cause of nosocomial and community infections (1) .It is usually present in the environment as well as in the microbiota of the superior respiratory tract and skin.MRSA is transmitted by direct contact, and patient-to-patient transmission is largely via the hands of health care providers (HCPs).Undetected MRSAcolonized HCPs represent a major source of the bacterium in hospitals, since it can be transmitted from these individuals to high-risk patients, which can limit the success of other control measures.Indeed, a case in which MRSA transmission was reduced after carrier HCPs were identifi ed and successfully decolonized has been described (2) .
Individuals with human immunodefi ciency virus/acquired immune defi ciency syndrome (HIV/AIDS) are at increased risk for MRSA colonization and infection mainly due to their high antibiotic use (1) and high rates of hospital readmission.These patients can also harbor community-acquired MRSA, which often produces Panton-Valentine leukocidin (PVL) (1) .However, the prevalence of MRSA colonization in this group is not well established.
We addressed this in the present study by investigating the rate of MRSA colonization in hospitalized AIDS patients and their HCPs.MRSA isolates were characterized genetically and phenotypically and compared to determine whether any transmission occurred between patients and/or HCPs.The isolates were also compared to isolate from 2009 that were collected from different patients at the same hospital.

Sample collection
MRSA samples from two colonization sites [nares (N) and saliva (S)] were obtained from HIV/AIDS patients on days 1 and 7 of hospitalization and from HCPs with whom they had contact, in two specifi c units of a large Brazilian public hospital with 600 beds, of which 24 are occupied almost exclusively by HIV patients.From August 2011 to May 2013, 317 individuals with HIV/AIDS were hospitalized, and 266 agreed to participate in the study along with 73 HCPs; staffs that were on leave were not included.Samples were collected using swabs and stored in Stuart agar until bacterial isolation and identifi cation.S. aureus was isolated on Mueller Hinton agar supplemented with 5% sheep blood and phenotypic identifi cation was carried out using the Vitek system (BioMérieux, Marcy l'Etoile, France).Methicillin resistance was detected using the AST-P585 card (BioMérieux) and broth dilution.Once an HCP was identifi ed as being colonized with MRSA, a decolonization protocol was carried out that included a chlorhexidine bath and application of 1ml silver sulfadiazine to the nares, which was repeated daily for 5 days.Additionally, 22 MRSA isolates from other patients from different wards at the same hospital collected between June and September 2009 that were archived by our group, and S. aureus strain N315 were included for determination of clonality.Isolates representing each pulsotype detected by pulsed-fi eld gel electrophoresis (PFGE) were randomly selected for typing by multilocus sequence typing (MLST) and Staphylococcus Cassette Chromosome (SCCmec) typing.The study protocol was approved by the Research Ethics Committee of the Ribeirão Preto School of Nursing (no.1304/2011).

Susceptibility profi ling
The minimum inhibitory concentration (MIC) was determined for oxacilin, vancomycin, teicoplanin, daptomycin, tigecyclin, linezolid, and quinupristin-dalfopristin by broth dilution, according to Clinical and Laboratory Standards Institute (CLSI) guidelines (3) .MIC 50 and MIC 90 were calculated and CLSI breakpoints were adopted for classification except in the case of tigecycline, for which the European Committee on Antimicrobial Susceptibility Testing (EUCAST) recommendation was followed (4) .

Staphylococcus aureus screening
Heterogeneous vancomycin-intermediate Staphylococcus aureus screening (hVISA) screening was performed as previously described (5) .If a countable number (one to 30) of colonies was observed within 48h of incubation at 37°C on brain-heart infusion agar containing 4μg/ml vancomycin, the strain was designated as a possible hVISA (5) .S. aureus Mu3 and Mu50 were used as hVISA and VISA control strains, respectively, and were kindly provided by Keiichi Hiramatsu and Teruyo Ito (Juntendo University, Tokyo, Japan).

Population analysis profi le
The population analysis profi le (PAP) (5) (6) to vancomycin or daptomycin was determined for samples that were positive for the hVISA screening or exhibited resistance to daptomycin after 48h of incubation (6) (7) (8) .PAP to daptomycin was used to identify heterogeneous daptomycin-resistant Staphylococcus aureus (hDRSA) strains (8) .

Panton-Valentine leukocidin gene and hemolysis analysis
The PVL gene lukSF was amplifi ed by PCR as previously described (9) .Mueller Hinton agar supplemented with 5% sheep blood was used to assess hemolytic activity.
PFGE was carried out after DNA digestion with SmaI (11) .Data were analyzed with Bionumerics v.7.1 software (Applied Maths NV, Belgium) (12) using the unweighted pair-group method with arithmetic mean based on Dice coefficients, where optimization and tolerance were set to 0.5% and 1.25%, respectively.A similarity coeffi cient of 80% was selected to describe patterns representative of each pulsotype, which were further characterized by MSLT (13) .Sequence types were identifi ed using the MLST database (14) .To limit redundancy, duplicate isolates from the same patient with identical SCCmec and pulsotype were considered as the same strain and included only once in the analysis.
Staphylococcus aureus was cultured from 101 (38%) tested individuals, and resistance to oxacillin was observed in 15 participants (5.6% of all participants or 14.8% of those colonized by S. aureus).A total of 13/15 HIV patients were found to be colonized by MRSA on the day of hospital admission, and only fi ve of these remained colonized on day 7. Additionally, two patients were found to be colonized only on day 7 of hospitalization.Only 3/73 (4.1%) of HCPs (P1, P2, and P3) were colonized by MRSA during the study.P1 was colonized on three different dates, despite having undergone decolonization procedures (15) .Therefore, a total of 29 MRSA isolates were characterized: 24 from patients and fi ve from HCPs (Figure 1).
Susceptibility analysis confi rmed β-lactam resistance in all putative MRSA isolates.Daptomycin resistance in isolate 80N was observed after a 48-h incubation; this was unexpected, since daptomycin is considered bactericidal.A population analysis of the isolate revealed a heterogeneous phenotype that included some daptomycin-resistant cells (Figure 2A).One isolate (176N) was identified as hVISA, which was confirmed by PAP (Figure 2B).According to CLSI breakpoints (3) , this isolate exhibited an intermediate level of resistance to teicoplanin after 24h of incubation (MIC = 16μg/ ml) and full resistance after 48h.However, it would be classifi ed as resistant after 24h based on EUCAST breakpoints (4) .MIC 50 and MIC 90 (μg/ml) of all antimicrobials were determined for each strain, and were as follows: 128/>256 for oxacilin, 4/4 for linezolid, 0.5/1 for teicoplanin, 0.125/0.5 for tigecycline, 0.5/0.5 for quinupristin/dalfopristin and daptomycin, and 1/2 for vancomycin.MICs for each isolate are shown in Figure 1.Only 2/29 isolates (6.9%) were PVL-positive (strains 199N and 273N), while 25 (86.2%) were fully hemolytic.Isolates from HIV patients showed considerable variation in terms of PFGE band patterns, which included eight pulsotypes (A-E, G, H, and J) in contrast to three pulsotypes (F, I, and K) among isolates from HCPs.