Pandemic clone USA300 in a Brazilian hospital: detection of an emergent lineage among methicillin-resistant Staphylococcus aureus isolates from bloodstream infections

Background Staphylococcus aureus is one of the leading causes of bloodstream infections (BSI) worldwide. In Brazil, the hospital-acquired methicillin-resistant S. aureus USA100/SCCmecII lineage replaced the previously well-established clones. However, the emergence of community-associated (CA) MRSA lineages among hospitalized patients is an increasing issue. Methods Consecutive S. aureus isolates recovered from BSI episodes of patients admitted between January 2016 and December 2018 in a Brazilian teaching hospital were tested for antimicrobial resistance, their genotypic features were characterized, and the clinical characteristics of the patients were evaluated. Results A total of 123 S. aureus isolates were recovered from 113 patients. All isolates were susceptible to linezolid, teicoplanin and vancomycin and 13.8% were not susceptible to daptomycin. Vancomycin MIC50 and MIC90 of 2 mg/L were found for both MRSA and MSSA isolates. The MRSA isolation rate was 30.1% (37/123), and 51.4% of them carried the SCCmec type II, followed by SCCmecIV (40.5%). Among the 37 MRSA isolates, the main lineages found were USA100/SCCmecII/ST5 and ST105 (53.7%) and USA800/ST5/SCCmecIV (18.9%). Surprisingly, six (16%) CA-MRSA isolates, belonging to USA300/ST8/SCCmecIVa that carried PVL genes and the ACME cassette type I, were detected. These six patients with USA300 BSI had severe comorbidities, including cancer, and most had a Charlson score ≥ 5; furthermore, they were in wards attended by the same health professionals. MRSA isolates were associated with hospital acquired infections (p = 0.02) in more elderly patients (p = 0.03) and those diagnosed with hematologic cancer (p = 0.04). Among patients diagnosed with MRSA BSI, 19 (54.3%) died. Conclusions The pandemic MRSA USA300 was detected for the first time in the Brazilian teaching hospital under study, and its cross-transmission most probably occurred between patients with BSI. This lineage may already be circulating among other Brazilian hospitals, which highlights the importance of carrying out surveillance programs to fight multidrug resistant and hypervirulent isolates.

Page 2 of 10 Augusto et al. Antimicrobial Resistance & Infection Control (2022) 11:114 Background Staphylococcus aureus is one of the most common causes of bloodstream infections (BSI) worldwide [1,2]. The presence of an indwelling device or a skin and soft tissue infection (SSTI) associated to hospital (HA) or community (CA) BSI are the most common sources of this pathogen [3]. S. aureus BSIs, especially those caused by MRSA, cause higher rates of morbidity and mortality and the costs associated to such infections are burdensome, especially in middle and low-income countries [4]. In addition, the presence of methicillin-resistant S. aureus (MRSA) as the cause of BSI represents a therapeutic challenge in health care institutions [5]. However, there has been a reduction in the number of MRSA isolated from BSI worldwide [1,6]. Diekema and coworkers [1] showed that S. aureus was the main Gram-positive bacteria isolated from BSI in a 20-year surveillance period, between 1997 and 2016, involving 45 nations, including those in Latin America. The authors showed that there was a decline in the number of MRSA isolates over these years and that daptomycin resistance among S. aureus isolates remained rare (< 0.1%); furthermore, during that period there was no trend toward an increase in the vancomycin MIC. Over recent years, the substitution of well-established MRSA lineages has been described in hospital environments [7,8]. In Brazilian hospitals, the previously predominant MRSA lineage, known as the Brazilian Endemic Clone (BEC/ST239) has been fully replaced by the New York/Japan lineage among BSI isolates (USA100/ST5 or ST105) [9][10][11]. Lately, our group also detected the emergence of USA1100/ST30, a CA-MRSA lineage in Brazilian hospital settings [8,10]. Curiously, in Brazil, the pandemic clone USA300/ST8/SCCmecIVa and the USA300-Latin American Variant (LV)/SCCmecIVc-e lineages have remained rare [11]. These related lineages present different SCCmec subtypes and the USA300-LV lacks the operon ACME [12]. However, these lineages could carry the Panton-Valentine leukocidin genes and are considered hypervirulent, since once they are present in the environment; they can spread easily, leading to infections with high mortality rates [13].
Molecular studies have helped the epidemiological surveillance of S. aureus BSI, highlighting the constant clonal change that has been taking place in health institutions and the dissemination of MRSA clones that can lead to higher rates of morbidity and mortality [14]. Moreover, constant monitoring of S. aureus BSI can also help surveillance programs in containing the spread of their resistance and virulence. In recent years, our group conducted a surveillance of S. aureus BSI in a University Hospital of Rio de Janeiro, Brazil, to identify the emergence of new clones or new resistance traits that may alter the epidemiology and treatment of these infections. This study aims to continue this surveillance through the characterization of S. aureus isolates from BSI over a three-year period in relation to antimicrobial resistance and virulence associated with different clones and correlate these results with clinical data of the patients.

Clinical isolates and setting
This prospective cohort study evaluated the phenotypic and molecular profiles of S. aureus isolates recovered from consecutive episodes of BSI occurring in adult patients (> 18 years old), who were admitted to the Hospital Universitário Clementino Fraga Filho (HUCFF) between January 2016 and December 2018. HUCFF is a tertiary public teaching hospital in Rio de Janeiro, Brazil, which currently has 300 active beds.
The first isolate of an episode of S. aureus BSI were analyzed, with subsequent documentation of blood cultures, clinical improvement, and anti-staphylococcal therapy. Demographic data, classification of the BSI episode [15], treatment and length of stay were collected from each patient.
All blood cultures were processed using the BacT/ ALERT system (BioMerieux, Durham, NC, USA). Bacterial identification was performed by the automated VITEK2 ® system (BioMerieux) and confirmed by MALDI-TOF MS (Matrix Assisted Laser Desorption Ionization/ Time of Flight Mass Spectrometry) (Bruker Daltonics, Billerica, MA, USA) [16].

Conclusions:
The pandemic MRSA USA300 was detected for the first time in the Brazilian teaching hospital under study, and its cross-transmission most probably occurred between patients with BSI. This lineage may already be circulating among other Brazilian hospitals, which highlights the importance of carrying out surveillance programs to fight multidrug resistant and hypervirulent isolates. oxacillin, teicoplanin and vancomycin (Sigma-Aldrich Chemical Company, St Louis, MO, USA) were determined by the broth microdilution (BMD) method, using fresh cation-adjusted Muller-Hinton broth (CAMHB). Daptomycin received a supplement of calcium (50 μg/ mL) [17]. The ceftaroline MICs were determined by the gradient diffusion method (Etest ® , BioMérieux) for all MRSA isolates. S. aureus ATCC 25923 and ATCC 29213 were used as controls for the disk-diffusion and BMD tests, respectively. The SCCmec typing [18] and subtyping [19] for isolates resistant to oxacillin by the disk-diffusion method were assessed by multiplex polymerase chain reaction (PCR), using bacterial DNA obtained using the QIAamp DNA Mini Kit (QIAGEN, Hilden, Germany). The Staphylococcus strains used as positive controls are described in Salgueiro et al. [20].

Phenotypic tests for screening of hVISA isolates
All isolates presenting a MIC ≥ 2 mg/L for vancomycin were screened for the hVISA (heteroresistant vancomycin-intermediate S. aureus) phenotype [21,22]. Plates with BHI agar (Becton Dickson, Franklin Lakes, NJ, USA), containing 3 and 4 mg/L of vancomycin were swabbed with 0.5 McFarland standard suspension (10 8 CFU/mL) of S. aureus and incubated at 35 to 37 °C for 24 h. Reduced susceptibility to vancomycin was defined as the growth of one or more colony-forming units (CFU) at either of the two concentrations. BHI agar plates with 4 mg/L of vancomycin and supplemented with 16 g/L of pancreatic casein digestion (Merck), were also inoculated with four 10μL spots from a 0.5 McFarland inoculum [22]. An isolate was considered to have reduced susceptibility to vancomycin if at least one spot had two or more CFUs. S. aureus ATCC 29,213 (MSSA) and Mu3 (hVISA) were used as controls [23].

Genotypic profile of MRSA isolates
All MRSA isolates were typed by pulsed-field gel electrophoresis (PFGE) after digestion of whole cell DNA with SmaI in a CHEF-DRIII system (Bio-Rad, Richmond, CA, USA), as previously described [24]. The PFGE fingerprints were compared by the unweighted pair-group method with arithmetic mean clustering analysis, applying the Dice correlation coefficient. Isolates with four or fewer bands of difference or minimum of 80% similarity were classified as belonging to the same genotype [25]. The clonal lineages were defined by comparison with national [10] and international clones [26]. Representative isolates from each genotype, identified in PFGE, underwent Multilocus Sequence Typing (MLST) [27].

Detection of virulence genes
The detection of PVL genes was carried out in all MRSA isolates using PCR [28]. Isolates belonging to the USA300/ST8/SCCmecIV lineage also underwent ACME typing [29]. The Staphylococcus strains used as positive controls are described in Schuenck et al. [28].

Statistical analysis
Categorical variables were compared using the chisquare or Fisher exact tests, and continuous variables were compared using the Wilcoxon test. A p value < 0.05 was considered to be statistically significant. All analyses were performed using SPSS 21.0 for Windows (SPSS, Inc).

Baseline characteristics of patients presenting S. aureus BSI
During the study period, 113 patients developed S. aureus BSI. Ten of them presented two BSI episodes. Thus, a total of 123 S. aureus isolates were recovered. Most of the patients (53.1%; 60/113), who were men, presented at least one comorbidity (89.4%) and developed central line-associated BSI (55.8%) or secondary bacteremia due to SSTI (24.8%). MRSA was associated to more elderly patients (p = 0.03), hematologic cancer (p = 0.04) and hospital-associated BSI (p = 0.02). A total of 50 (44.2%) patients died, and among those diagnosed with MRSA BSI, 54.3% died (Table 1).

Antimicrobial susceptibility, hVISA screening and SCCmec typing
According to the disk-diffusion method, 30.1% (37/123) of S. aureus isolates were resistant to cefoxitin and were characterized as MRSA. Overall, 51.4% (19/37) of these isolates carried the SCCmecII lineage, followed by SCCmecIV (15/37; 40.5%), SCCmecIII (2/37; 5.4%) and SCCmecV (1/37; 2.7%). Among the SCCmecIV isolates those that were subtyped as SCCmecIVa belonged to USA300 lineage. Only the SCCmecIII isolates were resistant to TMP/STX. Table 2 shows the MIC results. Both MRSA and MSSA isolates presented MIC 50 and MIC 90 = 2 mg/L for vancomycin. This MIC value was found in 90% of the SCCmecII isolates and in 53.3% of those that carried the SCCmecIV. The ceftaroline MICs in the MRSA isolates ranged from 0.25 to 0.75 mg/L. The oxacillin MICs, MIC 50 and MIC 90 , were 256 and ≥ 256 mg/L, respectively in the MRSA isolates, and 1 and 2 mg/L among the MSSA isolates. All isolates were susceptible to linezolid, teicoplanin and vancomycin. Seventeen (13.8%) S. aureus isolates were non-susceptible to daptomycin, and among them, five (29.4%) were MRSA isolates.  All the 72 (58.5%) isolates that presented a vancomycin MIC = 2 mg/L were screened for the hVISA phenotype. However, no hVISA phenotype was found.

Characteristics of the USA300 isolates
All the USA300/ST8/SCCmecIVa isolates (1963,1967,1982,1988,2013, and 2020) carried both PVL virulence genes and ACME cassette type I (Table 3). These isolates were recovered from different patients, who had developed hospital or healthcare-associated BSIs. All these individuals had severe comorbidities and all, but one patient (isolate number 2020), had a central line in situ when diagnosed with BSI. Four patients presented a Charlson score ≥ 5. Most of these patients were undergoing cancer treatment, although not at the same unit. Figure 2 shows the timeline distribution of the six patients with BSI caused by the USA300 according to their hospital ward. A direct nosocomial transmission could not be clearly established, although some patients were admitted to the hospital only a few days apart, and there was a ≥ 90% similarity between the isolates (Fig. 1). However, the wards occupied by these patients are physically close to each other (along the same corridor), and the same healthcare workers and cleaning staff attended these wards.

Discussion
In this study, we characterized consecutive S. aureus isolates from BSI in a teaching hospital in Rio de Janeiro for three years and the emergence of the pandemic CA-MRSA USA300/ST8/SCCmecIVa lineage was detected for the first time in this hospital. The patients with USA300 BSI were in wards that were attended by the same health professionals and, therefore, we hypothesized a cross-transmission of these isolates occurred in the hospital. Although other hospitals may not have the same epidemiological situation seen here, these results may reflect a picture of the circulating strains of S. aureus that cause BSI in Brazil.
Although the CA-MRSA lineages are classically related to SSTI, they have also been widely implicated in HA-BSI [30,31]. According to Kourtis et al. [32] CA-MRSA infections provide a reservoir that contributes to the incidence of healthcare-associated diseases and fuels transmission both within and outside healthcare settings. Our group has been carrying out a surveillance of S. aureus BSI in this hospital (HUCFF) for years and, although we have described the presence of community S. aureus lineages in this hospital since 2009 [8,10,[33][34][35] we had not yet seen the presence and spread of USA300/ST8/SCCme-cIVa. Some Brazilian studies have shown the presence of CA-MRSA isolates in hospitals recovered from different clinical sources, including occasionally USA300/ ST8/SCCmecIV isolates [36][37][38][39]. However, these studies did not characterize the isolates of this lineage accurately. Furthermore, different from what has been described in Latin American countries about the occurrence of the USA300-LV lineage [11], here, the isolates that belonged to USA300/ST8 carried the PVL genes, the SCCmec type IVa, as well as the ACME-I cassette, and were characterized as the USA300/ST8 North American lineage. Patients with USA300 BSI presented severe comorbidities and most had a Charlson score ≥ 5 and cancer. Curiously, among the patients with MRSA BSI this pathology was most frequently associated to those presenting USA300 (p < 0.05) (data not shown). Although we were not able to demonstrate a time-space relationship among all of them, a close genetic similarity of the isolates leads us to suggest a possible horizontal transmission associated with this clonal lineage. Furthermore, even if all the patients involved did not occupy the same ward at the same time, there would be a possible sharing of unscreened patients, which may be the focus of the outbreak, pointing out the need for constant surveillance of this pandemic clonal lineage within this hospital.
The present study shows a great diversity of MRSA clonal lineages causing BSI which was also observed in previous studies by our group in this teaching hospital [8,9]. Here, the USA100 lineage is the main MRSA clone (51.4%). These data confirm the replacement of the prior prevalent BEC clone, which now represents only 5.4% of MRSA isolates found. Bride et al. [38], in a study conducted in a hospital in the southeast of Brazil characterized S. aureus isolates from various clinical sources and the USA100/ST5 lineage was prevalent among HA-infections. However, in our study, most of the USA100 isolates belonged to ST105 (84.2%), a SLV from ST5. Although this is the first report of this lineage in our hospital, it has already been detected in S. aureus isolates from BSI in other Brazilian hospitals [8,9]. Viana and coworkers recently evaluated MRSA isolates from different hospitals in Rio de Janeiro, between 2014 and 2017 and observed the emergence of this SCCmecII/ST105 lineage, which has presented multidrug resistance characteristics and   11:114 seems to be associated with increased evasion when exposed to monocytic cells [40]. The successful spread of S. aureus isolates is commonly attributed to their antimicrobial resistance profile, but virulence may also play a crucial role in colonization and infection [41]. In the present study, the presence of PVL genes was evaluated in all MRSA isolates and only the CA-MRSA lineages related to USA300/ST8 and USA1100/ST30 were positive. Besides, USA300 presented the two virulence determinants investigated, PVL genes and ACME-I. Some authors have reported an association between virulence and mortality among patients with S. aureus BSI, indicating that the transmission of a virulent pathogen to an already debilitated patient can worsen his clinical condition [7]. However, more studies are needed to better understand the relevance of virulence determinants in S. aureus bloodstream infections.
Among Latin American countries, the MRSA isolates accounted for almost half of the S. aureus BSI described over the last few years [1,11]. In Brazil, the MRSA isolation rates vary according to the region. For example, a study conducted at a tertiary oncology care center in Rio de Janeiro detected 26.3% of methicillin resistance among S. aureus isolates recovered from BSI [42]. On the other hand, Primo and coworkers [4] showed a MRSA isolation rate of 58.3% among S. aureus BSI in a teaching hospital located in the central-west region of Brazil. The MRSA isolation rate in HUCFF in Rio de Janeiro has remained around 30% in BSI caused by S. aureus, since 2011, as previously described by our group [10], similarly to the findings of the present study, which reinforces the need for constant surveillance directed towards this pathogen.
A structured management in diagnosis and treatment of S. aureus BSI, which differs accordingly to methicillinresistance, is crucial for an optimal outcome [43]. Vancomycin is the first-line treatment for MRSA BSI worldwide [5], although non-susceptible vancomycin isolates have been already reported [21,44]. In the present study, no hVISA/VISA (Vancomycin-intermediate S. aureus) or VRSA (vancomycin-resistant S. aureus) isolates were detected. However, 58.5% of the isolates presented a high vancomycin MIC, and 76% (28/37) of them were MRSA isolates. The clinical efficacy of this antimicrobial can be critically affected by a MIC = 2 mg/L [21]. Thus, daptomycin has become an important alternative to treat MRSA BSI [45]. However, 13.5% of MRSA isolates in the present study were characterized as non-susceptible to daptomycin. Although daptomycin resistance among S. aureus remains rare worldwide (< 0.1%) [1], Silva et al. [46] also found isolates non-susceptible to daptomycin (4.7%) in a Brazilian study involving 128 clinical S. aureus isolates. We have already described BSI caused by non-susceptible daptomycin S. aureus isolates (MIC of 2 and 4 mg/L) among MRSA-VISA isolates in the hospital of the present study [2]. This data is of great concern because of its impact on the patient's outcome.
In a previous study, da Silva et al. suggested that comorbidities can contribute to a greater susceptibility for acquiring bacterial infections, such as MRSA BSI [47], in elderly patients. We also verified that MRSA BSI was associated with older patients (p = 0.03). In fact, in our study most patients (101/113; 89.4%) presented at least one comorbidity, which may indicate a close relationship with age. Although the global prevalence of MRSA BSI in patients with cancer is relatively low [48], in this study this pathogen was detected in 46% of the patients under study. Even though this study demonstrated that elderly and cancer patients were the most affected, we concluded that the high mortality rates found may not be due to MRSA infection, but to the severe health condition of the patients evaluated.
Some limitations of this study include a lack of national epidemiological data on S. aureus BSI, which could make comparative analyses in relation to clinical and microbiological aspects related to this type of infection difficult. Furthermore, we only characterized BSI S. aureus isolates, and we did not assess the colonized patients to confirm the hospital spread of USA300/ST8. Although we found S aureus isolates from other lineages disseminated in the hospital that seemed to be endemic, such as the USA100 lineage, the unusual presence of the pandemic and hypervirulent USA300/ST8 clone must be highlighted. This is because of its ability to establish itself  11:114 in hospital settings as has been well described in North American studies; highlighting the fact that this lineage could also become endemic in Brazilian hospitals.

Conclusions
The emergence of MRSA isolates of the USA300/ST8/ SCCmecIVa lineage was reported in the Brazilian teaching hospital under study. This possibly occurred due to an in-hospital spread among patients with BSI. Therefore, we conclude that this pandemic and hypervirulent lineage may already be circulating among Brazilian hospitals and could be associated to future hospital outbreaks.