Advertisement

Surveillance of multi-drug resistance phenotypes in Staphylococcus aureus in Japan and correlation with whole-genome sequence findings

  • Y. Hosaka
    Correspondence
    Corresponding authors. Address: Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, 4-2-1 Aobacho, Higashimurayama, Tokyo 189-0002, Japan. Tel.: +81 42 391 8211.
    Affiliations
    Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, Tokyo, Japan
    Search for articles by this author
  • K. Yahara
    Correspondence
    Corresponding authors. Address: Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, 4-2-1 Aobacho, Higashimurayama, Tokyo 189-0002, Japan. Tel.: +81 42 391 8211.
    Affiliations
    Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, Tokyo, Japan
    Search for articles by this author
  • A. Clark
    Affiliations
    WHO Collaborating Centre for Surveillance of Antimicrobial Resistance, Brigham and Women's Hospital, Boston, MA, USA
    Search for articles by this author
  • H. Kitagawa
    Affiliations
    Department of Infectious Diseases, Hiroshima University Hospital, Hiroshima, Japan

    Department of Surgery, Hiroshima University Graduate School of Medicine, Hiroshima, Japan
    Search for articles by this author
  • J. Hisatsune
    Affiliations
    Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, Tokyo, Japan
    Search for articles by this author
  • M. Sugai
    Affiliations
    Antimicrobial Resistance Research Center, National Institute of Infectious Diseases, Tokyo, Japan
    Search for articles by this author
  • K. Shibayama
    Affiliations
    Department of Bacteriology, Nagoya University Graduate School of Medicine, Aichi, Japan
    Search for articles by this author
  • J. Stelling
    Affiliations
    WHO Collaborating Centre for Surveillance of Antimicrobial Resistance, Brigham and Women's Hospital, Boston, MA, USA

    Harvard Medical School, Boston, MA, USA
    Search for articles by this author
Open AccessPublished:February 21, 2022DOI:https://doi.org/10.1016/j.jhin.2022.02.011

      Summary

      Background

      Antimicrobial resistance in Staphylococcus aureus imposes a high disease burden. Both phenotypic and genotypic monitoring are key to understanding and containing emerging resistant strains.

      Aim

      Phenotypic monitoring of emerging resistance in S. aureus and correlation of priority strain phenotypes with whole-genome sequencing (WGS) findings.

      Methods

      Antimicrobial susceptibility test results of >40,000 isolates from 213 participating hospitals from 2011 to 2019 were exported from the national Japan Nosocomial Infections Surveillance (JANIS) database. Longitudinal and geographic distribution and prevalence of distinct multi-drug resistance phenotypes (‘resistance profiles’) of S. aureus were examined among hospitals and prefectures. We further conducted a genome sequence analysis of strains with specific resistance profiles of concern.

      Findings

      The overall prevalence of meticillin-resistant S. aureus (MRSA) decreased from 40.3% to 35.1% from 2011 to 2019. However, among dozens of S. aureus resistance profiles, only one profile of a type of MRSA, exhibited a statistically significant increase in inpatient frequency, exceeding 10% during the nine years. This MRSA profile showed resistance to oxacillin, erythromycin and levofloxacin. Analysis of WGS results of S. aureus isolates with this phenotype revealed that most belonged to clonal complex 8, and all carried SCCmec IV, typical of community-acquired MRSA.

      Conclusion

      Tracking distinct resistance profiles deepened our understanding of the overall decrease in MRSA and led to recognition of the emergence of a new resistance phenotype. This study provides a model for future epidemiological research on antimicrobial resistance correlating multi-drug resistance phenotypes with selective genome sequencing, which can be applied to other bacterial species.

      Keywords

      Introduction

      Antimicrobial resistance (AMR) is a global threat because infection with resistant bacteria potentially has severe consequences, including prolonged illness, prolonged hospital stay, increased medical costs and increased mortality [
      World Health Organization
      Global action plan on antimicrobial resistance.
      ]. After defining AMR as one of the three most important public health threats of the 21st century in 2014, the World Health Organization (WHO) endorsed the adoption of the WHO Global Action Plan on AMR in 2015 [
      World Health Organization
      Global action plan on antimicrobial resistance.
      ,
      World Health Organization
      Antimicrobial resistance: global report on surveillance.
      ]. In this plan, AMR surveillance plays a vital role in the assessment of the burden of AMR and effective actions are supported for AMR containment, with action required at the local, national, and global level [
      World Health Organization
      Global antimicrobial resistance surveillance system.
      ].
      Drug-resistant Staphylococcus aureus, including meticillin-resistant S. aureus (MRSA), is among the most important AMR bacteria in the WHO AMR global surveillance report owing to its high virulence and a dearth of effective drugs for resistant strains [
      World Health Organization
      Antimicrobial resistance: global report on surveillance.
      ,
      • Tong S.Y.C.
      • Davis J.S.
      • Eichenberger E.
      • Holland T.L.
      • Fowler Jr., V.G.
      Staphylococcus aureus infections: epidemiology, pathophysiology, clinical manifestations, and management.
      ]. Regardless of the recent decline in the proportion of invasive MRSA infections in the US and many European countries [
      • Klein E.Y.
      • Mojica N.
      • Jiang W.
      • Cosgrove S.E.
      • Septimus E.
      • Morgan D.J.
      • et al.
      Trends in methicillin-resistant Staphylococcus aureus hospitalizations in the United States, 2010–2014.
      ,
      European Centre for Disease Prevention and Control (ECDC)
      Surveillance of antimicrobial resistance in Europe (2017).
      ], the mortality rate from S. aureus bacteraemia remains higher than that of AIDS, tuberculosis, and viral hepatitis, and increases if meticillin resistance is present [
      • van Hal S.J.
      • Jensen S.O.
      • Vaska V.L.
      • Espedido B.A.
      • Paterson D.L.
      • Gosbell I.B.
      Predictors of mortality in Staphylococcus aureus bacteremia.
      ]. A nationwide cohort study with 1184 sepsis patients in 59 intensive care units in Japan revealed that in-hospital mortality of MRSA sepsis patients was the highest among all causative pathogens, including meticillin-susceptible S. aureus (MSSA) [
      • Umemura Y.
      • Ogura H.
      • Takuma K.
      • Fujishima S.
      • Abe T.
      • Kushimoto S.
      • et al.
      Current spectrum of causative pathogens in sepsis: a prospective nationwide cohort study in Japan.
      ]. In another Japanese study using the data from 1133 acute care hospitals across Japan, the total estimated burden of MRSA in all 1584 acute care hospitals was US $2 billion (3.41% of total hospitalization costs), 4.34 million days (3.02% of total length of stay) and 14.3 thousand deaths (3.62% of total mortality) [
      • Uematsu H.
      • Yamashita K.
      • Kunisawa S.
      • Fushimi K.
      • Imanaka Y.
      Estimating the disease burden of methicillin-resistant Staphylococcus aureus in Japan: Retrospective database study of Japanese hospitals.
      ]. While the number of healthcare-associated MRSA (HA-MRSA) cases comprising the majority of isolated MRSA cases has declined over time, so-called community-acquired MRSA (CA-MRSA) has been rising since its emergence in the mid-1990s [
      • Crum N.F.
      The emergence of severe, community-acquired methicillin-resistant Staphylococcus aureus infections.
      ,
      • David M.Z.
      • Daum R.S.
      Community-associated methicillin-resistant Staphylococcus aureus: epidemiology and clinical consequences of an emerging epidemic.
      ,
      • Laupland K.B.
      Incidence of bloodstream infection: a review of population-based studies.
      ] (see David et al. [
      • David M.Z.
      • Daum R.S.
      Community-associated methicillin-resistant Staphylococcus aureus: epidemiology and clinical consequences of an emerging epidemic.
      ] for the definition of HA- and CA-MRSA).
      Recognizing that routine molecular typing of S. aureus is not a current reality, the use of multi-drug resistance phenotypes, as defined by results from a core set of priority antimicrobials, can facilitate local and national phenotype tracking of emerging threats. The Japan Nosocomial Infections Surveillance (JANIS) is a national surveillance system coordinated by the National Institute of Infectious Diseases (NIID) and covering the world's largest number of hospitals. The JANIS has been comprehensively collecting routine bacteriological test results from both symptomatic and asymptomatic patients through the voluntary co-operation of the clinical laboratories of participating hospitals since 2000 [
      • Tsutsui A.
      • Suzuki S.
      Japan nosocomial infections surveillance (JANIS): a model of sustainable national antimicrobial resistance surveillance based on hospital diagnostic microbiology laboratories.
      ,
      • Tsutsui A.
      • Yahara K.
      • Clark A.
      • Fujimoto K.
      • Kawakami S.
      • Chikumi H.
      • et al.
      Automated detection of outbreaks of antimicrobial-resistant bacteria in Japan.
      ]. This database includes the results of species identification and antimicrobial susceptibility testing of all routine specimens and all bacteria analysed in each hospital [
      • Hirabayashi A.
      • Yahara K.
      • Kajihara T.
      • Sugai M.
      • Shibayama K.
      Geographical distribution of Enterobacterales with a carbapenemase IMP-6 phenotype and its association with antimicrobial use: an analysis using comprehensive national surveillance data on antimicrobial resistance.
      ]. As of January 2021, 2283 hospitals contribute approximately monthly data to this database. In 2019, the JANIS database included results from almost six million isolates. The JANIS database has been used to address public health concerns, including the spread of AMR, by providing regional and national AMR summary reports with facility and prefectural benchmarking data. This is required for the development, implementation, and evaluation of evidence-based AMR control efforts [

      WHONET Developed by the WHO collaborating Centre for surveillance of antimicrobial resistance at the Brigham and Women’s hospital in Boston, USA. Available at: http://www.whonet.org/[last accessed November 2021].

      ]. The substantial nationwide data collected over 20 years is a valuable resource for exploring temporal trends and evolving geographical distributions of S. aureus resistance profiles, and also for recognizing and tracking novel and priority strains of public health concern.
      To take full advantage of the national surveillance system and create a basis for future epidemiological research, we analysed the comprehensive national AMR surveillance data. Phenotypic trends were explored in resistant profiles of multi-drug-resistant S. aureus that experienced statistically significant changes between 2011 to 2019, and strains with specific resistance profiles of concern were analysed by whole-genome sequencing (WGS) analysis.

      Methods

      Data sources and selection of a core set of antibiotics and hospitals

      All inpatient and outpatient data fields for all specimens collected in Japan between January 2011 and December 2019 were extracted from the JANIS database, which stores both culture-positive and culture-negative test diagnostic results with all antimicrobial susceptibility testing results. The extracted data were converted to WHONET format using a ‘JANIS to WHONET data parser’ and imported into WHONET, a free software programme developed and supported by the WHO Collaboration Centre for Surveillance of Antimicrobial Resistance since 1989. WHONET mainly focuses on the analysis of antimicrobial susceptibility results and supports local and national surveillance programmes in over 130 countries [

      WHONET Developed by the WHO collaborating Centre for surveillance of antimicrobial resistance at the Brigham and Women’s hospital in Boston, USA. Available at: http://www.whonet.org/[last accessed November 2021].

      ]. Because antimicrobial susceptibility test (AST) practices between facilities were not identical, and not all antibiotics were tested on every isolate, we sought to ascertain a subset of antimicrobials that were relatively consistently tested across hospitals and over the project time period. This subset of antimicrobials was used to construct ‘resistance profiles’ that could be compared between facilities and over time. We identified a core set of six antimicrobials that suitably met this a criterion, i.e., all the S. aureus isolates were tested with each antimicrobial in more than 90% of hospitals. The six antimicrobials included gentamicin (GEN), erythromycin (ERY), clindamycin (CLI), minocycline (MNO), levofloxacin (LVX) and oxacillin (OXA). OXA is used for the recognition of MRSA, although we recognized that practices for reporting or suppressing OXA results for MSSA and MRSA varied between facilities.
      The breakpoints of a subset of antimicrobials in the Clinical and Laboratory Standards Institute (CLSI) guideline did not change during the study period. Furthermore, we used the same CLSI guideline to categorize each isolate based on its AST result.
      After identifying the subset of antimicrobials, we selected hospitals where AST results for all six drugs were consistently available from 2011 to 2019. We excluded hospitals where all the S. aureus isolates were meticillin-resistant as a likely artifact of local practices for the suppression of OXA results among MSSA isolates. Hospitals with fewer than 10 S. aureus isolates per year on average were also excluded because of the small sample size.

      Data tabulation

      After removal of isolates missing results for one or more of the core antimicrobials, we used a Java programme developed at NIID to extract aggregated annual data for the number of S. aureus isolates stratified by 26 + 1 resistance profiles (i.e., combinations of susceptible and non-susceptible results of the six drugs, as well as a summation of all other resistance profiles, including intermediate results) from 2011 to 2019. De-duplication was conducted according to standard JANIS criteria to remove repeated isolates of the same species isolated from a patient within a 30 day period, regardless of specimen type, but considering the AMR phenotype [
      • Kajihara T.
      • Yahara K.
      • Stelling J.
      • Eremin S.R.
      • Tornimbene B.
      • Thamlikitkul V.
      • et al.
      Comparison of de-duplication methods used by WHO Global Antimicrobial Resistance Surveillance System (GLASS) and Japan Nosocomial Infections Surveillance (JANIS) in the surveillance of antimicrobial resistance.
      ]. The de-duplication procedure selects and counts isolates with significantly different drug susceptibilities as different isolates, even if they were isolated within 30 days from the same patient. A subsequent isolate was selected and counted if it showed discordance from susceptible to resistant (or vice versa), or a four-fold or greater change in the minimum inhibitory concentration value for a specific antimicrobial when compared with a previous isolate from the same patient within the 30-day period. We used an in-house Perl script to tabulate the aggregated data for each prefecture, as in our previous study [
      • Kajihara T.
      • Yahara K.
      • Stelling J.
      • Eremin S.R.
      • Tornimbene B.
      • Thamlikitkul V.
      • et al.
      Comparison of de-duplication methods used by WHO Global Antimicrobial Resistance Surveillance System (GLASS) and Japan Nosocomial Infections Surveillance (JANIS) in the surveillance of antimicrobial resistance.
      ]. We then calculated the annual proportions of each multi-drug resistance profile as the number of isolates divided by the total number of isolates subjected to antimicrobial susceptibility testing with the six core antimicrobials. We tabulated these proportions separately for the inpatient and outpatient settings.

      Longitudinal and geographical analyses of the resistance profiles

      Longitudinal and geographical distributions of frequency of specific resistance profiles in each prefecture from 2011 to 2019 were visualized using JMP (JMP 14, SAS Institute Inc., Cary, NC, USA) and R version 3.5.2. The chi-squared test for trend was performed to test whether there was a trend between the frequency of a specific resistance profile and year using R version 3.5.2.

      WGS analysis

      To investigate the genomic background of specific resistance profiles, we utilized a historical dataset of genome sequences and AST results from two strain collections: (1) all S. aureus bacteraemia strains reported from 2013 to 2017 at Hiroshima University Hospital, which has participated in JANIS since 2000; and (2) 183 representative S. aureus strains across 26 clonal complexes (CCs) isolated from 19 prefectures in Japan. The 183 representative strains (Japan clone library) were historically selected based on pulsed-field gel electrophoresis of various clinically isolated strains from the different prefectures, followed by two-dimensional clustering using BioNumerics GelCompar software (Applied Maths NV. Belgium) and selection of strains representative of the clusters. AST results of the strains are not yet available. As in a previous study [
      • Yahara K.
      • Nakayama S-i
      • Shimuta K.
      • Lee K-i
      • Morita M.
      • Kawahata T.
      • et al.
      Genomic surveillance of Neisseria gonorrhoeae to investigate the distribution and evolution of antimicrobial-resistance determinants and lineages.
      ], we first conducted pairwise genome alignment between the reference strain JP140 and one of the other strains using progressiveMauve [
      • Darling A.E.
      • Mau B.
      • Perna N.T.
      progressiveMauve: multiple genome alignment with gene gain, loss and rearrangement.
      ]. We then combined the alignments into a multiple whole-genome alignment, in which each position corresponded to that of the reference genome. We constructed a maximum-likelihood tree using PhyML [
      • Guindon S.
      • Dufayard J.-F.
      • Lefort V.
      • Anisimova M.
      • Hordijk W.
      • Gascuel O.
      New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0.
      ] from a genome alignment containing 206,600 core single nucleotide polymorphisms. The tree with information on specific CCs, resistance profiles, and staphylococcal chromosomal cassette mec (SCCmec) of each strain was visualized using Phandango [
      • Hadfield J.
      • Croucher N.J.
      • Goater R.J.
      • Abudahab K.
      • Aanensen D.M.
      • Harris S.R.
      Phandango: an interactive viewer for bacterial population genomics.
      ].

      Ethics

      Patient identifiers were de-identified by each hospital before data were submitted to JANIS. The anonymous data stored in the JANIS database were exported and analysed following approval by the Ministry of Health, Labor and Welfare (approval number 0624–5) according to Article 32 of the Statistics Act.

      Results

      Longitudinal trends of MRSA with specific resistance profiles

      We identified 213 hospitals in the JANIS database that submitted AST results for all six drugs (OXA, GEN, ERY, CLI, MNO and LVX) consistently between 2011 and 2019. The hospitals reported more than 10 S. aureus isolates on average per year and they did not show 100% MRSA prevalence in any year. Among these hospitals, the number of de-duplicated S. aureus isolates gradually declined from 53,508 to 42,645 (20.0%) from 2011 to 2019 in inpatient settings, while the number increased from 22,067 to 26,259 (19.0%) during the same period in outpatient settings (Supplementary Figure S1). The percentage of MRSA among S. aureus isolates in inpatient settings decreased from 45.8% to 38.8% during the nine-year period, but in outpatient settings, the percentage of MRSA fluctuated around 27.0% over the same period. Overall, the percentage of MRSA declined from 40.3% in 2011 to 35.1% in 2019.
      Among the 26 possible multi-drug resistance profile combinations, only two (both MRSA) showed a decrease (one profile) or increase (one profile) over 10.0% in frequency during the study period. The profile showing a decrease included strains resistant to all six antimicrobials (6-drugs = OXA GEN ERY CLI MNO LVX), whereas the profile showing an increase included strains resistant to the three specific antimicrobials (3-drugs = OXA ERY LVX). Figure 1 shows the number and proportion of these two priority resistance profiles, as well as the total of all other MRSA, from 2011 to 2019 in the inpatient and outpatient settings.
      Figure 1
      Figure 1Annual trend in number and proportion of meticillin-resistant Staphylococcus aureus (MRSA) isolates resistant to 6-drugs (oxacillin (OXA), gentamicin (GEN), erythromycin (ERY), clindamycin (CLI), minocycline (MNO), levofloxacin (LVX)), 3-drugs (OXA, ERY, LVX), and all other MRSA isolates between 2011 and 2019. Each bar indicates the number of MRSA isolates resistant to 6-drugs (red), 3-drugs (blue), and all other MRSA (green). Each line indicates the proportion of MRSA isolates resistant to 6-drugs (red), 3-drugs (blue), and all other MRSA (green) using the total number of isolates subjected to the 6-drugs antimicrobial susceptibility testing (26 +1 resistance profiles) as the denominator. The number and proportion of MSSA are not shown in this figure. Graphs are shown separately for inpatient (upper) and outpatient (lower) samples.
      Strains resistant to the 6-drugs showed a decreasing trend (P<2.2e-16, chi-squared trend test), whereas strains resistant to the 3-drugs steadily increased over the nine years (P<2.2e-16, chi-squared trend test) (Figure 1). There was not a large difference between inpatient and outpatient trends, although the extent of change in inpatient settings was larger than that in outpatient settings. Specifically, the proportion of strains resistant to the 6-drugs in inpatient settings exhibited the largest decline among all the priority resistance profiles, from nearly 18.0% to approximately 4.0% (red line), throughout the nine-year period. In contrast, the frequency of strains resistant to the 3-drugs was low in 2011 in both inpatient and outpatient settings (around 2.0%), but it substantially increased to 14.7% (inpatient) and 11.7% (outpatient) in 2019 (blue line). The frequency of all other MRSA gradually decreased in both inpatient and outpatient settings with a larger change in inpatient settings over the study period (green line).

      Geographical features

      To explore the geographical distribution of specific resistance profiles between 2011 and 2019, the annual proportion of strains resistant to the 6-drugs (a) and those resistant to the 3-drugs (b) was calculated for each prefecture in inpatient settings (Figure 2). The proportion of strains resistant to the 6-drugs decreased during this period, both the median (from 15% to 5.0%) and the range (from 38.0% to 15.0%). In contrast, the median (0.8–14.0%) and range (16.0–39.0%) of the proportion of strains resistant to the 3-drugs increased over the same time period.
      Figure 2
      Figure 2Annual trends in proportion of strains resistant to 6-drugs (oxacillin, gentamicin, erythromycin, clindamycin, minocycline, levofloxacin) and 3-drugs (oxacillin, erythromycin, levofloxacin) from 2011 to 2019. Each dot corresponds to each prefecture. Each value was calculated as the number of strains resistant to the (a) 6-drugs and (b) 3-drugs divided by the total number of isolates subject to the 6-drugs antimicrobial susceptibility testing (26 +1 resistance profiles) in each prefecture in each year. In the box plot, the bottom and top of the box indicate the 25th and 75th percentile, respectively; the horizontal line in the box indicates the median; and the top outliers are above the 75th percentile + 1.5× interquartile range.
      The geographical distribution of the 3-drug MRSA phenotype is shown on prefectural maps in Figure 3. In 2011, the proportion of strains resistant to the 3-drugs was smaller than 10.0% in almost all prefectures except for Shizuoka (15.9%) in the southwest of Japan (Figure 3(a)). The average proportion of the 3-drug profiles increased during the nine-year period from 1.7% in 2011 to 14.6% in 2019. The proportion was higher than 20.0% in 10 of 47 prefectures in 2019. The highest proportions of strains resistant to the 3-drugs in 2019 were in Tochigi (39.3%), Tottori (34.5%), and Yamanashi (31.0%) (Figure 3 (b1), (b2) and (b3), respectively). Tochigi and Yamanashi are both located in the region adjacent to the Tokyo metropolitan area, whereas Tottori is not.
      Figure 3
      Figure 3Prefectural maps of the proportion of Staphylococcus aureus resistant to 3-drugs in each prefecture in 2011 and 2019. Prefectures where isolates resistant to the 3-drugs were not isolated are coloured in white, and others are coloured in red using five different shades (with bin size 10%). In the 2011 map (left), the (a) Shizuoka prefecture (15.2%), where the proportion of isolates resistant to the 3-drugs was highest, is indicated by an arrow. In the 2019 map (right), the top three prefectures where the proportion of strains resistant to the 3-drugs was more than 30% are indicated by arrows: (b1) Tochigi (39.3%), (b2) Tottori (34.5%), and (b3) Yamanashi (31.0%).

      Molecular phylogenetic analysis of the MRSA strains resistant to the 3-drugs

      To explore the genetic features of strains with notable resistance profiles, we retrieved historical WGS and AST phenotypic data from all MRSA blood isolates (67 isolates) from a single university hospital participating in JANIS between 2013 and 2017. The strains resistant to the 3-drugs accounted for 9.0% (six of 67 during the 5 years). A phylogenetic tree of core genomes was constructed from these six strains, as well as representative S. aureus strains from another strain collection across 26 CCs isolated from 19 prefectures in Japan. The tree showed that most of the strains resistant to the 3-drugs (orange in Figure 4) belonged to CC8 (blue in Figure 4). Classification of staphylococcal cassette chromosome mec (SCCmec) showed that all of the strains resistant to the 3-drugs carried SCCmec IV, typical of CA-MRSA (Supplementary Figure S2). These results contrast with findings from strains resistant to the 6-drugs, almost all of which carried SCCmec II (Supplementary Figure S2), typical of HA-MRSA. Detailed information of CC, ST, SCCmec, and AST results of the strains resistant to the 3-drugs are shown in Supplementary Table S1.
      Figure 4
      Figure 4Phylogenetic relatedness of strains resistant to the 3-drugs. A phylogenetic tree was constructed of core-genomes of strains resistant to oxacillin (OXA) erythromycin (ERY) and levofloxacin (LVX) combined with those of other representative Staphylococcus aureus strains across 26 clonal complexes (CCs) isolated in 19 prefectures in Japan. Left: maximum-likelihood phylogeny. Right: heatmap. Column 1: CC5 and CC8 are coloured red and blue, respectively. Column 2, yellow indicates the strains resistant to OXA, ERY, and LVX, whereas purple indicates other representative S. aureus strains. Column 3, staphylococcal cassette chromosome mec (SCCmec) types I, Ⅱ, IV, and V are coloured light pink, dark pink, light blue, and dark blue, respectively. Scale bar indicates the number of substitutions per nucleotide site in the reference genome.

      Discussion

      The most notable finding of this study was the substantial decrease in the proportion of MRSA isolates resistant to all six core antimicrobials (OXA, GEN, ERY, CLI, MNO, LVX) and the increase in the proportion of those resistant to the 3-drugs (OXA, ERY, LVX) from 2011 to 2019. It is expected that the decline in the proportion of the 6-drugs resistant MRSA reflects the reduction of HA-MRSA in inpatient and outpatient settings. This is because the susceptibility of 6-drugs resistance MRSA to more than two non-β-lactam antibiotics, including CLI, GEN and ciprofloxacin, is most commonly seen in HA-MRSA typically carrying SCCmec type I, II, or III. In general, HA-MRSA isolates carry a relatively large SCCmec (type I, II or III), which typically harbours several antibiotic resistance genes, whereas CA-MRSA strains carry a smaller SCCmec (most commonly type IV or V) containing fewer resistance genes [
      • David M.Z.
      • Daum R.S.
      Community-associated methicillin-resistant Staphylococcus aureus: epidemiology and clinical consequences of an emerging epidemic.
      ,
      • David M.Z.
      • Glikman D.
      • Crawford S.E.
      • Peng J.
      • King K.J.
      • Hostetler M.A.
      • et al.
      What is community-associated methicillin-resistant Staphylococcus aureus?.
      ,
      • Hiramatsu K.
      • Katayama Y.
      • Yuzawa H.
      • Ito T.
      Molecular genetics of methicillin-resistant Staphylococcus aureus.
      ]. Several Japanese studies have shown that most SCCmec type II MRSA strains have lower susceptibility to CLI, LVX, ERY, GEN and MNO than that of isolates with SCCmec type IV, commonly identified in CA-MRSA [
      • Yanagihara K.
      • Araki N.
      • Watanabe S.
      • Kinebuchi T.
      • Kaku M.
      • Maesaki S.
      • et al.
      Antimicrobial susceptibility and molecular characteristics of 857 methicillin-resistant Staphylococcus aureus isolates from 16 medical centers in Japan (2008-2009): Nationwide survey of community-acquired and nosocomial MRSA.
      ,
      • Yamaguchi T.
      • Nakamura I.
      • Chiba K.
      • Matsumoto T.
      Epidemiological and microbiological analysis of community-associated methicillin-resistant staphylococcus aureus strains isolated from a Japanese Hospital.
      ,
      • Yamaguchi T.
      • Okamura S.
      • Miura Y.
      • Koyama S.
      • Yanagisawa H.
      • Matsumoto T.
      Molecular characterization of community-associated methicillin-resistant Staphylococcus aureus isolated from skin and pus samples of outpatients in Japan.
      ,
      • Mitsumoto-Kaseida F.
      • Murata M.
      • Toyoda K.
      • Morokuma Y.
      • Kiyosuke M.
      • Kang D.
      • et al.
      Clinical and pathogenic features of SCCmec type II and IV methicillin-resistant Staphylococcus aureus in Japan.
      ].
      A surge in CA-MRSA is becoming a serious problem in the control of MRSA infections, regardless of a large decline in HA-MRSA infections [
      • David M.Z.
      • Daum R.S.
      Community-associated methicillin-resistant Staphylococcus aureus: epidemiology and clinical consequences of an emerging epidemic.
      ,
      • Chuang Y.-Y.
      • Huang Y.-C.
      Molecular epidemiology of community-associated meticillin-resistant Staphylococcus aureus in Asia.
      ]. Regarding the increase in frequency of MRSA resistant to the 3-drugs, our genome sequencing analysis revealed that all of these strains harboured SCCmec type IV (Figure 4), which indicates an increase in CA-MRSA among MRSA strains in both inpatient and outpatient settings.
      In previous studies on SCCmec type IV strains (typically CA-MRSA) in Japan, the susceptibility rate of LVX was higher than that of ERY (45.0–60.0% versus 10.0–30.0%). Thus, the majority of type IV MRSA strains were susceptible to LVX [
      • Yamaguchi T.
      • Okamura S.
      • Miura Y.
      • Koyama S.
      • Yanagisawa H.
      • Matsumoto T.
      Molecular characterization of community-associated methicillin-resistant Staphylococcus aureus isolated from skin and pus samples of outpatients in Japan.
      ,
      • Mitsumoto-Kaseida F.
      • Murata M.
      • Toyoda K.
      • Morokuma Y.
      • Kiyosuke M.
      • Kang D.
      • et al.
      Clinical and pathogenic features of SCCmec type II and IV methicillin-resistant Staphylococcus aureus in Japan.
      ,
      • Chuang Y.-Y.
      • Huang Y.-C.
      Molecular epidemiology of community-associated meticillin-resistant Staphylococcus aureus in Asia.
      ,
      • Ryu S.
      • Cowling B.J.
      • Wu P.
      • Olesen S.
      • Fraser C.
      • Sun D.S.
      • et al.
      Case-based surveillance of antimicrobial resistance with full susceptibility profiles.
      ]. This is somewhat inconsistent with our observation of a large increase in the 3-drugs phenotype identified in this study, which were non-susceptible to LVX. The strains with the 3-drugs phenotype were not USA300-type usually resistant to LVX, but had nonsynonymous substitutions in the quinolone resistance-determining regions (S80F in grlA genes and S84L in gyrA). The discordance in these findings might be due to an increase in the proportion of type IV MRSA strains resistant to LVX from 2015 to 2019, as earlier studies were conducted before 2015.
      The exploration of geographical features of specific resistance profiles first confirmed a downward trend of strains resistant to the 6-drugs (Figure 2a) and an upward trend of those resistant to the 3-drugs (Figure 2b) among prefectures across the country Prefectural maps (Figure 3). Then, visualization of the prefectural map (Figure 3) revealed the three prefectures with the highest proportion of strains resistant to the 3-drugs among all the prefectures in 2019 were not limited to a specific area. Moreover, the proportion of strains resistant to the 3-drugs tended to be higher in eastern than western Japan (the median was 16.1% and 9.7%, respectively). Further studies are warranted to explore additional factors (e.g., clinical practices) underlying such geographical differences in the distribution of strains resistant to the 3-drugs.
      We utilized comprehensive national AMR surveillance data by examining combinations of resistance to several key antibiotics (‘resistance profiles’). The value of reporting and analysing such ‘full susceptibility profiles’ rather than each of the antibiotics of interest separately, was recently highlighted [
      • Ryu S.
      • Cowling B.J.
      • Wu P.
      • Olesen S.
      • Fraser C.
      • Sun D.S.
      • et al.
      Case-based surveillance of antimicrobial resistance with full susceptibility profiles.
      ]. Our study supports this approach and provides a model of how analysis of such full resistance profiles can be useful and effective in recognizing novel threats. This approach can also be used to monitor evolving microbial populations using routine phenotypic data as a practical, inexpensive proxy for genotypic characterization.
      Our study had several limitations. First, differentiation between MSSA and MRSA was based on phenotypic findings with OXA and was not confirmed by molecular typing and confirmation of the presence of the mecA gene. This could lead to over- or underestimation of MRSA prevalence and the proportion of drug-resistant MRSA. However, differentiation based on phenotypes is more practical in real settings than genetic analysis. Second, the criteria for collecting patient samples for culture and AST were not identical among facilities. This is unavoidable in the JANIS system, where hospitals have provided specimen results voluntarily utilizing routine clinical samples. This could also affect the prevalence and proportion of resistance profiles tabulated from the JANIS data. This study at least excluded hospitals with 100% MRSA as a likely artifact of local practices for the suppression of OXA results among MSSA isolates. Third, although approximately 80.0% of hospitals with more than 500 beds across Japan participate in JANIS, hospitals with fewer than 200 beds account for less than 14% of hospitals participating in JANIS. Thus, data from larger hospitals can have a greater impact on the findings than those from smaller hospitals [
      • Tsutsui A.
      • Suzuki S.
      Japan nosocomial infections surveillance (JANIS): a model of sustainable national antimicrobial resistance surveillance based on hospital diagnostic microbiology laboratories.
      ]. Finally, the isolates from Hiroshima University hospital used in molecular phylogenetic analysis might not be nationally representative of the strains resistant to the 3-drugs.
      Despite these limitations, to our knowledge, this is the first study to explore longitudinal and geographic distributions of specific AMR S. aureus resistance profiles over nine years at the national level by leveraging the comprehensive national phenotypic AMR surveillance database. Furthermore, this study is the first to conduct genome sequence analysis of strains showing characteristic susceptibility patterns identified in the resistance profile analysis, and demonstrates how WGS can complement routine microbiological tests in hospital laboratories [
      • Argimón S.
      • Masim M.A.L.
      • Gayeta J.M.
      • Lagrada M.L.
      • Macaranas P.K.V.
      • Cohen V.
      • et al.
      Integrating whole-genome sequencing within the National Antimicrobial Resistance Surveillance Program in the Philippines.
      ]. Identifying and characterizing key resistance profiles deepened our understanding of the contributors to the overall decrease of MRSA in Japan and revealed the emergence in many centres of a new resistance phenotype with features of CA-MRSA that we further characterized using WGS. This study provides a model for future epidemiological research on AMR based on resistance profiles and WGS, which can be applied to other bacterial species.

      Acknowledgements

      The authors are grateful to all the participating hospitals for their collaboration and for contributing their data to JANIS.

      Conflict of interest statement

      The authors have no conflicts of interest to disclose.

      Funding sources

      This work was supported by the Research Program on Emerging and Re-emerging Infectious Diseases from the Japan Agency for Medical Research and Development (AMED) (grant number JP21fk0108604 ).

      Author contributions

      Y.H. and K.Y. designed the study. A.C. and J.S. developed and maintained the WHONET-JANIS interoperability features. H.K. and J.H. provided data for genome sequencing. K.Y. curated and tabulated the data using software. K.S and M.S acquired the funding. Y.H. investigated the study. J.S, M.S. and K.S supervised the study. Y.H. and K.Y. validated and visualized the data. Y.H. wrote the original manuscript. Y.H., K.Y., J.S., H.K., J.H., M.S. and K.S. reviewed and edited the manuscript.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • World Health Organization
        Global action plan on antimicrobial resistance.
        2015 (Available at:) ([last accessed November 2021])
        • World Health Organization
        Antimicrobial resistance: global report on surveillance.
        2014 (Available at:) ([last accessed November 2021])
        • World Health Organization
        Global antimicrobial resistance surveillance system.
        2015 (Available at:) ([last accessed November 2021])
        • Tong S.Y.C.
        • Davis J.S.
        • Eichenberger E.
        • Holland T.L.
        • Fowler Jr., V.G.
        Staphylococcus aureus infections: epidemiology, pathophysiology, clinical manifestations, and management.
        Clin Microbiol Rev. 2015; 28: 603-661
        • Klein E.Y.
        • Mojica N.
        • Jiang W.
        • Cosgrove S.E.
        • Septimus E.
        • Morgan D.J.
        • et al.
        Trends in methicillin-resistant Staphylococcus aureus hospitalizations in the United States, 2010–2014.
        Clin Infect Dis. 2017; 65: 1921-1923
        • European Centre for Disease Prevention and Control (ECDC)
        Surveillance of antimicrobial resistance in Europe (2017).
        2018 (Available at:) ([last accessed November 2021])
        • van Hal S.J.
        • Jensen S.O.
        • Vaska V.L.
        • Espedido B.A.
        • Paterson D.L.
        • Gosbell I.B.
        Predictors of mortality in Staphylococcus aureus bacteremia.
        Clin Microbiol Rev. 2012; 25: 362-386
        • Umemura Y.
        • Ogura H.
        • Takuma K.
        • Fujishima S.
        • Abe T.
        • Kushimoto S.
        • et al.
        Current spectrum of causative pathogens in sepsis: a prospective nationwide cohort study in Japan.
        Int J Infect Dis. 2021; 103: 343-351
        • Uematsu H.
        • Yamashita K.
        • Kunisawa S.
        • Fushimi K.
        • Imanaka Y.
        Estimating the disease burden of methicillin-resistant Staphylococcus aureus in Japan: Retrospective database study of Japanese hospitals.
        PLoS One. 2017; 12e0179767
        • Crum N.F.
        The emergence of severe, community-acquired methicillin-resistant Staphylococcus aureus infections.
        Scand J Infect Dis. 2005; 37: 651-656
        • David M.Z.
        • Daum R.S.
        Community-associated methicillin-resistant Staphylococcus aureus: epidemiology and clinical consequences of an emerging epidemic.
        Clin Microbiol Rev. 2010; 23: 616-687
        • Laupland K.B.
        Incidence of bloodstream infection: a review of population-based studies.
        Clin Microbiol Infect. 2013; 19: 492-500
        • Tsutsui A.
        • Suzuki S.
        Japan nosocomial infections surveillance (JANIS): a model of sustainable national antimicrobial resistance surveillance based on hospital diagnostic microbiology laboratories.
        BMC Health Serv Res. 2018; 18: 1-10
        • Tsutsui A.
        • Yahara K.
        • Clark A.
        • Fujimoto K.
        • Kawakami S.
        • Chikumi H.
        • et al.
        Automated detection of outbreaks of antimicrobial-resistant bacteria in Japan.
        J Hosp Infect. 2019; 102: 226-233
        • Hirabayashi A.
        • Yahara K.
        • Kajihara T.
        • Sugai M.
        • Shibayama K.
        Geographical distribution of Enterobacterales with a carbapenemase IMP-6 phenotype and its association with antimicrobial use: an analysis using comprehensive national surveillance data on antimicrobial resistance.
        PLoS One. 2020; 15e0243630
      1. WHONET Developed by the WHO collaborating Centre for surveillance of antimicrobial resistance at the Brigham and Women’s hospital in Boston, USA. Available at: http://www.whonet.org/[last accessed November 2021].

        • Kajihara T.
        • Yahara K.
        • Stelling J.
        • Eremin S.R.
        • Tornimbene B.
        • Thamlikitkul V.
        • et al.
        Comparison of de-duplication methods used by WHO Global Antimicrobial Resistance Surveillance System (GLASS) and Japan Nosocomial Infections Surveillance (JANIS) in the surveillance of antimicrobial resistance.
        PLoS One. 2020; 15e0228234
        • Yahara K.
        • Nakayama S-i
        • Shimuta K.
        • Lee K-i
        • Morita M.
        • Kawahata T.
        • et al.
        Genomic surveillance of Neisseria gonorrhoeae to investigate the distribution and evolution of antimicrobial-resistance determinants and lineages.
        Microb Genomics. 2018; 4 (e000205)
        • Darling A.E.
        • Mau B.
        • Perna N.T.
        progressiveMauve: multiple genome alignment with gene gain, loss and rearrangement.
        PLoS One. 2010; 5: e11147
        • Guindon S.
        • Dufayard J.-F.
        • Lefort V.
        • Anisimova M.
        • Hordijk W.
        • Gascuel O.
        New algorithms and methods to estimate maximum-likelihood phylogenies: assessing the performance of PhyML 3.0.
        Syst Biol. 2010; 59: 307-321
        • Hadfield J.
        • Croucher N.J.
        • Goater R.J.
        • Abudahab K.
        • Aanensen D.M.
        • Harris S.R.
        Phandango: an interactive viewer for bacterial population genomics.
        Bioinformatics. 2018; 34: 292-293
        • David M.Z.
        • Glikman D.
        • Crawford S.E.
        • Peng J.
        • King K.J.
        • Hostetler M.A.
        • et al.
        What is community-associated methicillin-resistant Staphylococcus aureus?.
        J Infect Dis. 2008; 197: 1235-1243
        • Hiramatsu K.
        • Katayama Y.
        • Yuzawa H.
        • Ito T.
        Molecular genetics of methicillin-resistant Staphylococcus aureus.
        Int J Med Microbiol. 2002; 292: 67-74
        • Yanagihara K.
        • Araki N.
        • Watanabe S.
        • Kinebuchi T.
        • Kaku M.
        • Maesaki S.
        • et al.
        Antimicrobial susceptibility and molecular characteristics of 857 methicillin-resistant Staphylococcus aureus isolates from 16 medical centers in Japan (2008-2009): Nationwide survey of community-acquired and nosocomial MRSA.
        Diagn Microbiol Infect Dis. 2012; 72: 253-257
        • Yamaguchi T.
        • Nakamura I.
        • Chiba K.
        • Matsumoto T.
        Epidemiological and microbiological analysis of community-associated methicillin-resistant staphylococcus aureus strains isolated from a Japanese Hospital.
        Jpn J Infect Dis. 2012; 65: 175-178
        • Yamaguchi T.
        • Okamura S.
        • Miura Y.
        • Koyama S.
        • Yanagisawa H.
        • Matsumoto T.
        Molecular characterization of community-associated methicillin-resistant Staphylococcus aureus isolated from skin and pus samples of outpatients in Japan.
        Microb Drug Resist. 2015; 21: 441-447
        • Mitsumoto-Kaseida F.
        • Murata M.
        • Toyoda K.
        • Morokuma Y.
        • Kiyosuke M.
        • Kang D.
        • et al.
        Clinical and pathogenic features of SCCmec type II and IV methicillin-resistant Staphylococcus aureus in Japan.
        J Infect Chemother. 2017; 23: 90-95
        • Chuang Y.-Y.
        • Huang Y.-C.
        Molecular epidemiology of community-associated meticillin-resistant Staphylococcus aureus in Asia.
        Lancet Infect Dis. 2013; 13: 698-708
        • Ryu S.
        • Cowling B.J.
        • Wu P.
        • Olesen S.
        • Fraser C.
        • Sun D.S.
        • et al.
        Case-based surveillance of antimicrobial resistance with full susceptibility profiles.
        JAC-Antimicrobial Resistance. 2019; 1: dlz070
        • Argimón S.
        • Masim M.A.L.
        • Gayeta J.M.
        • Lagrada M.L.
        • Macaranas P.K.V.
        • Cohen V.
        • et al.
        Integrating whole-genome sequencing within the National Antimicrobial Resistance Surveillance Program in the Philippines.
        Nat Commun. 2020; 11: 2719