INTRODUCTION
Infections due to methicillin-resistant staphylococcal strains (MRS) present increased risk of treatment failure, and for some time glycopeptides, such as vancomycin and teicoplanin, have been the only therapeutic option, which justifies the crescent use of this class of drugs1),(2. Some antimicrobials that exhibit in vitro activity against Staphylococcus aureus remain active against strains resistant to glycopeptides such as rifampicin and fusidic acid3. In recent years, a limited number of new antimicrobials have been developed. Among them, the 5thgeneration cephalosporins, ceftaroline and ceftobiprole have been shown to be effective against MRSA isolates4. Other antimicrobials not belonging to the group of beta-lactams and with activity against these micro-organisms, including linezolid, daptomycin and tigecycline, have been available since the beginning of the 21st century and are widely employed in clinical practice5.
The indiscriminate use of vancomycin therapy to treat hospital infections has allowed for the emergence of S. aureus isolates and other staphylococcal species with reduced sensitivity to vancomycin and other glycopeptides6, such as vancomycin/glycopeptide intermediate resistant S. aureus (VISA/GISA) and vancomycin/glycopeptide resistant S. aureus (VRSA/GRSA)3. Studies have reported an increased number of glycopeptide-resistant Staphylococcus epidermidis (GRSE) worldwide7.
Isolates of S. aureus, especially coagulase-negative Staphylococci (CoNS) resistant to methicillin/oxacillin isolates, with reduced susceptibility to glycopeptides have been reported in Japan, the United States (US), Europe, and Asia since the end of the 80s3. In 1997, two major categories of vancomycin-resistant S. aureus had been defined: (1) vancomycin-resistant S. aureus (VRSA), carrying the vanA gene, mediating high-level resistance (MIC ≥ 16µg/ml) and (2) vancomycin-intermediate S. aureus (VISA) isolates with low-level resistance (MIC ≥ 4 to < 16µg/ml) by cell wall thickening8.
In Brazil, staphylococcal strains with reduced susceptibility to glycopeptides were reported in hospitals in São Paulo and Rio de Janeiro a decade ago9, and recently S. aureus containing the vanA and vanB genes were described in Brazil10),(11.
Colonization of diverse body sites by CoNS, and transient colonization by S. aureus, can be a source of infection for immunocompromised patients12, highlighting microbiota colonization of health professionals as a source of dissemination of health care-associated infections (HAIs)13.
The transmission chain of antimicrobial resistant microorganisms in the hospital environment involves health professionals as potential source of transmission to patients, co-workers, family and community, emphasizing their importance in the context of the HAIs13),(14.
In Brazil, resistant phenotypic profiles of different Staphylococcus species, with reduced resistance to vancomycin, were detected in microbiota samples of asymptomatic carriers 12. Considering that Staphylococcusspecies are important pathogens associated with HAIs and there are few studies regarding vancomycin resistance in the colonizing microbiota of health professionals, the aims of this research were to identify the microorganisms colonizing the nasopharynx of health professionals and analyze the vancomycin resistance profile of these isolates by phenotypic and genetic methods.
METHODS
Study design and bacterial samples
An experimental-based study was carried out in April to December 2014 with staphylococcal isolates obtained in a previous study15, originated from nasopharyngeal secretions of healthcare professionals of three sectors of the University Hospital of Pernambuco, Brazil: Intensive Care Unit (ICU), Surgical Clinics, and Hemodialysis Service/Nephrology.
A total of 102 Staphylococcus strains-kept as frozen stock in Brain Heart-Infusion (BHI) broth supplemented with glycerol (20%) at -20°C and in nutrient agar slants at 4ºC-were placed in BHI broth, inoculated in 5% sheep blood agar, and incubated for 24-48h at 35ºC. Colonies with macroscopic characteristics of the genus Staphylococcus were Gram stained. Following confirmation by morphology and staining, colonies were submitted for identification using deoxyribonuclease (DNase), catalase and coagulase tests, and manitol fermentation. The identification of the staphylococcal species was made through the automated system VITEK 2. Based on these readings, profile identification was established and interpreted according to a specific algorithm. The result of the profile was compared with the database, generating the identification of the unknown organism.
Vancomycin susceptibility testing
The isolates were screened by vancomycin susceptibility test using BHI supplemented with 6µg/mL of vancomycin (BHI-V6). Inoculums were adjusted to 0.5 McFarland turbidity3.
Two methods were used to determine the MIC to vancomycin: broth microdilution and E-test (BIOMÉRIEUX), according to Clinical and Laboratory Standards Institute (CLSI) guidelines16. One clinical isolate of Enterococcus faecium harboring the vanA gene was used as a positive control17 and the Enterococcus faecalis (ATCC 29212) strain as the negative control. The E-test was performed using a suspension of 0.5 McFarland turbidity plated onto Müller-Hinton medium and incubated at 35ºC for 24h. The interpretation was performed according to manufacturer’s specifications and compared to CLSI cut off values3.
For determination of MIC by Broth Microdilution, it was determined manually in Mueller Hinton broth, according to the recommendations of the CLSI16. Assays for vancomycin were performed in medium supplemented with Ca2 + (50mg/L). Initial inoculums of bacteria (0.5 × 105 CFU/mL) were plated onto 96-well polypropylene plates, exposed to 8 dilutions (1μg/mL to 128μg/mL) of the tested compound, and incubated for 18h at 35°C. The minimum inhibitory concentration was taken as the lowest concentration of the compound in which no visible bacterial growth was observed. According to CLSI recommendations, the bacterial isolates were categorized as resistant or susceptible using interpretive criteria16.
Agreement between E-test and microdilution was defined as minimum inhibitory concentrations (MICs) that differed by ± 1-log2 dilutions or less. Categorical agreement was defined as test results within the same susceptibility. Errors were ranked as follows: very major error, false-susceptible result by the E-test; major error, false-resistant result produced by the E-test; and minor error, intermediate result by E-test method and a resistant or susceptible category for the reference method (microdilution test), according to CLSI guidelines19.
Antimicrobial susceptibility testing
The Staphylococcus isolates were tested by disc diffusion in Mueller-Hinton agar, according to the CLSI guidelines18, using the following antibiotics: penicillin (10U), gentamicin (10µg), clindamycin (2µg), sulfazotrim (1.25/23.75µg), ciprofloxacin (5µg), chloramphenicol (30µg), cefoxitin (30µg), erythromycin (15µg), and linezolid (5µg). After incubation for 24 hours at 35°C, the inhibition zones were measured using a caliper.
Concordance scale analysis of phenotypic vancomycin susceptibility tests
The agreement between the phenotypic tests to assess the vancomycin resistance profile were verified by the k (kappa) index20.
Molecular techniques
Deoxyribonucleic acid (DNA) was extracted as described by Oliveira21 from vancomycin-resistant isolates detected by the different phenotypic techniques. Subsequently, it was used in polymerase chain reaction (PCR) for amplification of the genes vanA and mecA. PCR was performed using the primers and conditions as previously described (forward-5´-TGAATAACATCGGCATTAC-3´ and reverse-5´-TTATTTAACGGGGAAATC-3´)22 and (P1 5´-GGTCCCATTAACTCTGAAG-3´ and P3 5´-AGTTCTGCAGTACCGGATTTGC-3´)23.
Sequencing of vanA gene
A positive PCR product for the vanA gene was purified by the Wizard® SV Gel kit and PCR Clean-Up System (Promega) according to the manufacturer’s protocol. Following, it was quantified by spectrophotometry using the software Chromas Lite 2.1.1, Basic Local Alignment Research Tool (BLAST), and Expert Protein Analysis System (ExPASy) algorithm. The analyzed sequences of vanA were deposited in GenBank with the following accession number: KT581638.
RESULTS
In this study, we analyzed 102 isolates collected from health professionals, 31.4% (32/102) S. aureus and 68.6% (70/102) CoNS isolates. Approximately 43.1% (44/102) of the isolates were from the Surgical/Infectious and Parasitic Diseases sector, of which 22.7% (10/44) were S. aureus and 77.3% (34/44) were CoNS. Isolates from the ICU represented 20.6% (21/102), of which 38% (8/21) were S. aureus and 62% (13/21) were CoNS.
Concerning the vancomycin susceptibility, seven out of 19 isolates were S. aureus and 12 CoNS, which grew at the vancomycin screening test at a concentration of 6µg/mL (BHI-V6). The susceptibility profile to other antimicrobials showed that 18 isolates presented resistance to more than three classes of antimicrobials and were considered multi-drug resistant (MDR) strains. It is worth noting that 16 were resistant to erythromycin. Of these, 11 were also resistant to clindamycin, indicating resistance to macrolides, lincosamide, and streptogramin-B.
All 19 resistant strains, previously detected by the screening test, were evaluated by E-test for quantitative determination of the MIC to vancomycin. The seven isolates of S. aureus presented the following MIC ranges: two isolates with MIC ≤ 2μg/mL (sensitive), three with MIC between 4 and 8μg/mL, and two with MIC > 256μg/mL. In addition, of the 12 CoNS, 10 isolates presented MIC ≤ 4μg/mL and two isolates MIC ≥ 128μg/mL.
According to the MIC values obtained by broth microdilution, of the seven S. aureus isolates, five were sensitive (MIC ≤ 2µg/mL) and two were resistant (MIC ≥ 16µg/mL). Regarding the 12 CoNS isolates, nine showed MIC ≤ 2µg/mL and three MIC ≥ 32µg/mL, as shown in Table 1.
TABLE 1: Susceptibility profile of the Staphylococcus isolates.
| Bacterial species | Susceptibility profile | Cefoxitin disc-diffusion | BHI-V6 | E-test (µg/mL) | Broth microdiluition (µg/mL) | Gene vanA | Gene mecA |
|---|---|---|---|---|---|---|---|
| S. aureus | PEN, CLI, CFO, ERY | R | + | 4 | 2 | – | – |
| S. aureus | PEN, CLI, GM, CFO, RIF, ERY, CIP | R | + | 1 | 1 | – | – |
| S. aureus | PEN, CHL, CFO, ERY | R | + | 4 | 2 | – | – |
| S. aureus | PEN, CLI, ERY, CIP | R | + | > 256 | 128 | – | + |
| S. aureus | PEN, CLI, CFO, LZD, RIF, ERY, CIP | R | + | > 256 | 32 | – | – |
| S. aureus | PEN, CIP | R | + | 2 | 2 | – | + |
| S. aureus | PEN, GM, CFO, ERY, CIP | R | + | 4 | 1 | – | – |
| CoNS | PEN, CLI, SUF, CHL, CFO, ERY | R | + | 128 | 128 | – | + |
| CoNS | SUF, CHL, CFO, CIP | R | + | 2 | 2 | – | + |
| CoNS | PEN, CLI, SUF, CFO, LZD, RIF, ERY, CIP | R | + | 4 | 2 | – | + |
| CoNS | CLI, CHL, GM, CFO, ERY, CIP | R | + | 2 | 2 | – | + |
| CoNS | CLI, GM, CFO, ERY, CIP | R | + | 4 | 2 | – | + |
| CoNS | PEN, CLI, CFO, ERY, CIP | R | + | 4 | 2 | – | + |
| CoNS | PEN, CLI, CFO, ERY, CIP | R | + | 2 | 2 | – | + |
| CoNS | PEN, SUF, CFO, ERY | R | + | > 256 | 32 | + | + |
| CoNS | PEN, CLI, SUF, GM, CFO, RIF, ERY, CIP | R | + | 2 | 2 | – | + |
| CoNS | CFO, LZD, CIP | R | + | 2 | 2 | – | – |
| CoNS | PEN, CLI, CHL, CFO, ERY, CIP | R | + | 2 | 2 | – | + |
| CoNS | PEN, CLI, GM, CFO, ERY, CIP | R | + | 4 | 2 | – | + |
BHIV6: brain heart infusion agar; vanA: gene; mecA: gene; S.: Staphylococcus; CoNS: coagulase-negative Staphylococci; PEN: penicillin; CLI: clindamycin; CFO: cefoxitin; ERY: erythromycin; GM: gentamicina; RIF: rifampicin; CIP: ciprofloxacin; CHL: chloramphenico; LZD: linezolide; SUF: sulfazotrim; R: resistant; +: growth; -: growth.
The kappa index coefficient of 0.96 was obtained by comparing the E-test to Broth Microdilution, indicating very good agreement between these two methods.
The agreement within 1 two-fold dilution between E-test and the broth microdilution reference method was 84%. Seven (36%) minor errors were found comparing E-test with microdilution in broth for vancomycin. There was no occurrence of major or very major errors. The categorical concordance was 93%.
Of the five isolates with MICs considered resistant, only one isolate (S. epidermidis) carried the vanA gene, which was confirmed by sequencing and showed 100% of similarity with ten sequences from Enterococcus deposited in GenBank (KT581638).
In parallel to obtaining vancomycin susceptibility, a cefoxitin susceptibility test was performed using the disc-diffusion technique, from which it was possible to observe that all the isolates were resistant to cefoxitin. After the disc-diffusion cefoxitin test, the presence of the mecA gene was investigated.
In order to genetically characterize these isolates, mecA gene detection was performed by PCR, and the gene was found in 13/19 (39.4%) isolates, including one isolate of S. aureus (MRSA) from the Nephrology/Hemodialysis Service sector. Of the 12 MRCoNS, five were obtained from the Nephrology/Hemodialysis Service sector, four from Surgical/Infectious and Parasitic Diseases sector, and three from ICU. The isolate that presented the gene vanA also exhibited the gene mecA. As for the remaining noncarrier isolates of the vanA gene, five contained the mecA gene (all MRCoNS) and 13 did not present this gene (6 MSCoNS and 7 MSSA).
DISCUSSION
In this study, we found a higher incidence of mecA gene in CoNS strains. Regarding the susceptibility profile analyses, the MRS strains were more resistant to multiple classes of antimicrobial agents than MRSA. Similar results were obtained by Costa4; however, Fadeyi25 described MDR in MRSA isolates colonizing the nasopharynx of health professionals in Nigeria.
The studies regarding the analysis of susceptibility to vancomycin started approximately three decades ago, when environmental strains of S. aureus were detected with reduced susceptibility (intermediate) to this drug26. Additionally, the emergence of hetero-VRSA strains occurred in the 80s after the introduction of vancomycin use for treatment of staphylococcal infections in Japan27. Several research studies related to the molecular analysis of heteroresistance vancomycin-intermediate S. aureus worldwide have been performed, following associations of these strains with persistent infections and treatment failure28.
Results of MIC to vancomycin evaluated by E-test and broth microdilution techniques demonstrated divergence. Three isolates of S. aureus showed MICs between 4 and 8μg/mL (VISA) with the microdilution method but no isolates presented similar MIC with E-test. In addition, this technique is a screening tool for heterogeneous VISA (hVISA) and VISA, but does not apply to vancomycin or teicoplanin, and the results obtained with this technique should not be reported as true MIC29.
In the present study, four resistant vancomycin isolates (two VRSA and two VRS) were detected by both the E-test and broth microdilution. There was reasonable correlation between these two methods. Using comparisons between Broth Microdilution and E-test MICs results for vancomycin, it was possible to observe that essential and categorical agreements presented at 84% and 93%, respectively. Yet, a minor error of 36% was detected: however, major error and very major error were 0%.
Genetic characterization these vancomycin resistant isolates was performed by PCR for the vanA gene. During the period that the present study was conducted, vancomycin resistance was not described in other genes in Brazil10. Among the isolates only one harbored the vanA gene, which was a specific isolate Staphylococcus epidermidisfrom HCW microbiota.
The presence of the vanA gene was not found in the four resistant vancomycin strains (VRSA and VRS), as a result, the precise genetic mechanism for vancomycin resistance in these staphylococcal strains awaits elucidation. The cell wall thickening has been reported for glycopeptide-resistant VRS and VRSA12),(17),(28.
The nasopharynx microbiota of health professionals harboring resistant strains to vancomycin have already been described in the literature13),(30),(31; however, the first Brazilian report of S. epidermidis, harboring the vanAgene, and colonizing a health professional from ICU occurred at an University Hospital in Recife, Brazil12. Breves11 reported the occurrence of one S. aureus isolate obtained from the hands of a health professional, which harbored resistance genes to vancomycin (vanB) and methicillin (mecA). On the other hand, there are many studies reporting methicillin-resistant staphylococcal isolates from nasopharyngeal of health professionals and microbiota of patients32–34.
Cases reports regarding microbiota colonization of vancomycin-resistant staphylococcal strains obtained from patients at clinics or hospitals are scarce. In India, two studies have reported patients harboring S. aureus with the vanA gene35),(36.
The occurrence of HAIs caused by Staphylococcus isolates carrying the vanA gene has been reported worldwide, associated with different patterns of infections, mainly in the US and in Brazil11),(22),(37–41),(24.
Some studies correlate methicillin resistance to vancomycin tolerance due to vancomycin treatment failures in cases of infections caused by methicillin-resistant microorganisms7),(42–44. In 2006, three strains were reported, two S. aureus and one CoNS, with resistance to both antibiotics vancomycin and methicillin45.
A potential emergence of vancomycin resistance may occur in hospitals in Brazil, as reported in recent studies9),(28. This suggests the need for constant monitoring of susceptibility patterns to vancomycin, application of molecular methods and heteroresistance detection, as well as adoption of control measures to avoid the spread of these strains in the hospital environment.
