Outbreak of Streptococcus pyogenes emm89 ST646 in a head and neck surgical oncology ward

ABSTRACT Streptococcus pyogenes causes a variety of human infections, and hospital outbreaks with this pathogen have also been reported. The purpose of this study is to describe the clinical characteristics of an outbreak of S. pyogenes involving 15 patients and four healthcare workers (HCWs), as well as the molecular characteristics of the causative isolates. The course and response to the outbreak were reviewed, and information on the characteristics of the patients was extracted retrospectively from the medical records. Whole-genome sequencing of the 16 causative isolates (14 from patients and two from HCWs) was also performed. All 15 patients were postoperative of head and neck cancer with tracheotomy, and 12 had invasive infections, primarily surgical site infections, all of which resolved without causing serious illness. All but the first case was detected more than 7 days after admission. S. pyogenes was detected in two patients after empiric antimicrobial administration was performed on all inpatients and HCWs, and the outbreak was finally contained in approximately 2 months. All isolates detected in patients and HCWs belonged to emm89/clade 3, a hypervirulent clone that has emerged worldwide and was classified as sequence type 646. These isolates had single nucleotide polymorphism (SNP) differences of zero to one, indicating clonal transmission. This study demonstrated an outbreak of S. pyogenes emm89/clade 3 in a ward of patients with head and neck cancer. The global emergence of hypervirulent isolates may increase the risk of outbreaks among high-risk patients. IMPORTANCE This study describes an outbreak of Streptococcus pyogenes that occurred in a ward caring for patients with head and neck cancer and tracheostomies. Many cases of invasive infections occurred in a short period, and extensive empiric antimicrobial administration on patients and healthcare workers was performed to control the outbreak. Whole-genome sequencing analysis of the causative strains confirmed that it was a monoclonal transmission of strains belonging to emm89/clade 3. The epidemiology and clinical characteristics of S. pyogenes infections have changed with the replacement of the prevalent clones worldwide. In the 1980s, there was a reemergence of S. pyogenes infections in high-income countries due to the spread of hypervirulent emm1 strains. emm89/clade 3 has recently been spreading worldwide and shares common features with emm1, including increased production of two toxins, NADase, and streptolysin O. The outbreak reported here may reflect the high spreading potential and virulence of emm89/clade 3.

sepsis with or without bacteremia.Mortality from S. pyogenes infections now occurs mainly in low-and middle-income countries, where malnutrition, poor sanitation, and overpopulation exist (1).S. pyogenes infections had diminished in high-income countries from around 1950 likely owing to the amelioration of the factors listed above.However, it reemerged in developed countries in the 1980s and has become a major health problem again (2).This reemergence is attributed to the development of a pandemic type, emm1, which has acquired hypervirulent traits through several horizontal gene transfer events (3,4).
emm typing is a method to genetically classify S. pyogenes based on the homology of genes encoding the M protein, the most important virulence factor, and more than 200 emm types have been reported thus far (5).The global distribution of emm types has changed over time.Although isolates of the same emm type share common genetic characteristics and are sometimes associated with specific diseases, genetic diversity can occur even among isolates of the same emm type.Invasive infections caused by emm89 isolates have been increasingly reported worldwide in recent years (6)(7)(8)(9).A recent meta-analysis of the global distribution of emm types among S. pyogenes clinical isolates showed that emm89 is one of the most frequently detected emm types in high-income countries, along with emm1, emm12, emm28, etc. (10).Interestingly, the increase in emm89 has been accompanied by the emergence and rapid domination of a novel clade within emm89, clade 3.
A similar situation was observed in Japan, where genetic analysis of S. pyogenes isolates associated with STSS isolated in 2013-2018 showed that emm89 accounted for 23.3% of all isolates and was the second most common type after emm1 (11).In addition, another study showed that 156 (96.9%) of 161 emm89 clinical isolates from 2011 to 2019 in Japan belonged to clade 3 (12).During the coronavirus disease 2019 (COVID- 19) pandemic, when the number of STSS cases in Japan was reduced, the proportion of emm89 strains remained stable, despite a decrease in the proportion of emm1 strains among the causative strains (13).
In this study, we describe an outbreak in a hospital ward involving postoperative patients with head and neck cancer and healthcare workers and present the results of genetic analysis supporting the clonal spread of emm89/clade 3 isolates.

Setting and design
This was a retrospective descriptive study of the clinical and molecular characteristics of an outbreak of S. pyogenes in the head and neck surgery ward of a 686-bed cancer hospital.The head and neck surgery ward consists of 10 single-patient rooms and nine four-patient rooms, and more than 90% of the patients admitted are diagnosed with head and neck cancer.Nearly all patients admitted to the hospital for elective surgery for head and neck cancer are admitted to this ward, and most postoperative outpatients are also admitted to this ward unless the main purpose of admission is cancer chemother apy.Inpatients of the ward who have recently undergone surgery and/or have tracheot omy receive wound and airway care from physicians every morning sequentially in the procedure room of the ward, which has two treatment units accompanied by tables equipped with instruments and disposable items for the procedure.The case definition of patients in this study was inpatients of the ward with S. pyogenes detected in culture tests from April through June 2019.The case definition of healthcare workers (HCWs) in this study was HCWs dedicated to the head and neck surgery ward (non-physicians) or department (physicians) who had S. pyogenes detected in culture tests or were positive for group A streptococci in rapid antigen tests of throat swabs in the same period.
Isolation of bacteria from blood samples was performed using the BACT/ALERT 3D system (bioMérieux Japan Ltd., Tokyo, Japan).Isolation of S. pyogenes from clinical specimens was performed using horse blood agar medium (Kyokuto Pharmaceutical Industrial Co., Ltd., Tokyo, Japan).Identification of S. pyogenes in the microbiology laboratory of the hospital was performed with MicroScan WalkAway (Beckman Coulter, Brea, CA).Antimicrobial susceptibility testing was performed by disk diffusion methods using Kirby-Bauer disks (Eiken Chemical Co., Ltd., Tokyo, Japan) and the results were interpreted according to the CLSI M100-S27 guidelines (14).

Clinical data collection
The following clinical information was collected from electronic medical records for patients meeting the case definition: age; gender; type of malignancy diagnosed; date of admission; date of surgery prior to S. pyogenes detection; date of collection and specimen type of the culture test from which S. pyogenes was detected; wound or airway procedure in the ward procedure room within 7 days prior to S. pyogenes detection; administration of antimicrobial agents active against S. pyogenes within 7 days prior to S. pyogenes detection; the presence of infection due to S. pyogenes; diagnosis of S. pyogenes infection according to CDC/NHSN criteria (15); cure of S. pyogenes infection within 30 days following S. pyogenes detection; ICU admission due to S. pyogenes infection; mortality within 90 days following S. pyogenes detection; and presence of tracheotomy.

Molecular analysis
Whole-genome sequencing of 14 S. pyogenes isolates from patients and two S. pyo genes isolates from HCWs was performed with MiSeq (Illumina, San Diego, CA).Library preparation for MiSeq sequencing was performed with Nextera XT DNA Library Prep Kit (Illumina).Libraries were sequenced on a MiSeq system for 600 cycles (300 bp paired-end reads).Raw reads generated by MiSeq were quality-trimmed with fastp (version 0.23.1) and assembled using SPAdes (version 3.13.1).
We performed a phylogenetic analysis of the representative strain of this outbreak (FUJ00398) and the emm89 strains isolated in Japan that have been reported in previous articles and registered in GenBank.A PubMed search was conducted using the search terms "emm89" and "Japan" to identify the relevant publications.Genomic sequencing data of the targeted strains was obtained from GenBank (accession number: DRA009110 and AP017629).Genomic sequence reads of the target strains and MGAS11027, MGAS23530, and MGAS27061 was aligned to the genomic sequence of the reference strain (MGAS23530) using the BWA with the "sw" option.We obtained the provisional core-genome sequences including homologous recombination sites and SNPs using SAMtools mpileup and VarScan mpileup2cns, both with default parameters.A phyloge netic tree of the provisional core-genome sequences of the strains was estimated using PhyML.Using this as the starting tree, we inferred homologous recombination sites, which should be excluded from the core genomes, using ClonalFrameML (20).SNPs were determined by excluding SNPs that were present in the homologous recombination sites from SNPs that were present in the provisional core genome.The phylogenetic tree based on SNPs present in the core-genome excluding homologous recombination sites was estimated by maximum likelihood estimation using axmlHPC-PTHREADS command and GTRGAMMA model of RAxML (version 8.2.12) (21).

Outbreak description
S. pyogenes was detected from blood cultures of two inpatients of the ward taken on 9 April 2019 (Table 1).In response, the results of the microbiology tests of all inpatients of the ward in the previous month were reviewed by the infection control team, and the detection of S. pyogenes from culture specimens of two other patients submitted on 3 April and 6 April was recognized.Subsequently, S. pyogenes was detected from culture specimens collected from three other patients in the same ward on 11, 14, and 16 April, which led to the initiation of the intervention by the hospital infection control team.Ward HCWs were educated by the infection control team on adherence to hand hygiene and standard precautions.Standard precautions included the use of gloves and gowns during contact with exudate and the use of masks and eye protection during anticipated droplet contamination, but universal masking during work was not required.
Due to the subsequent increase in cases of S. pyogenes infection, active surveillance culture of pharyngeal swabs was performed and empiric antimicrobial administration was initiated on 23 April for all inpatients in the ward except for those who have received antimicrobial treatment active against S. pyogenes within 7 days (Fig. 1a).Because of the sharp increase in cases, empirical antimicrobials were administered without waiting for the results of the culture tests.The antimicrobial agent empirically administered was cephalexin 1,000 mg/day orally for 10 days, and one patient with a history of β-lactam allergy received azithromycin 500 mg/day for 3 days as an alternative agent.Due to the limited supply of antimicrobials to the hospital during this period, cephalexin, which was available in the necessary doses, was used as an alternative to penicillin.Active surveillance culture of pharyngeal swabs and empiric antimicrobial administration was also performed for all 65 ward HCWs including 17 physicians and 34 nurses from 19 April through 30 April (Fig. 1b).As with the patients, cephalexin was prescribed as the empiric antimicrobial, and two HCWs with a history of β-lactam allergy received azithromycin.It was also confirmed that no HCWs had skin lesions suspicious of skin and soft tissue infection.Environmental cultures were collected from patient chairs, handrails, headrests, treatment tables, and light handles in the procedure room, and repeated environmen tal disinfection of the patient rooms and the procedure room was implemented with alcohol-containing wipes.
Finally, a total of 15 inpatients were identified with S. pyogenes between 3 April and 30 May in the ward (Table 1).All strains were susceptible to penicillin, ampicillin, erythromy cin, and levofloxacin.All S. pyogenes isolates from the patients were detected in clinical culture specimens submitted at the discretion of the treating clinicians, and none were detected through active surveillance cultures performed as a part of the outbreak response (Fig. 1a).Detection of S. pyogenes from patients peaked during the third week of the outbreak, and new detections temporarily ceased after implementation of extensive empiric antimicrobial administration for patients and HCWs, but there were new detections in clinical culture specimens from one patient each on 17 May and 30 May (Fig. 2).One of these two patients was admitted after above-mentioned empiric antimicrobial administration for inpatients and was on his 12th day of hospitalization, and the other patient was more than 30 days post-admission and had been negative for active surveillance culture and had received empiric antimicrobial administration.These incidences suggested the persistence of transmission even after extensive empiric antimicrobial administration.During this period, there were no cases of S. pyogenes detection from patients in the ward other than those with head and neck cancer.Of the 65 HCWs for whom surveillance cultures were taken, two nurses (3.1%) tested positive in the third week (Fig. 1 and 2).Although they were identified by active surveillance culture, they had symptoms of pharyngitis at the time the test results arrived.Another two nurses had negative surveillance cultures but positive rapid antigen tests for S. pyogenes at clinics where they had visited with pharyngitis symptoms on other days (Table 1; Fig. 1).All four nurses improved quickly after the administration of antimicrobial agents, and  The risk of transmission from patients in the same room was considered positive if a patient in the same room had shedding of S. pyogenes within 2 days of S. pyogenes detection.S. pyogenes shedding was defined as up to 24 hours after the initiation of effective antimicrobial therapy, provided that S. pyogenes carriage was documented by culture.d M, male; F, female; NA, not applicable.
they were suspended from work until the administration of antimicrobial agents was completed and their symptoms had improved.Environmental cultures collected from multiple locations in the procedure room were negative for S. pyogenes.The first S. pyogenes detection from an inpatient of the ward after 30 May was on 26 August.The number of S. pyogenes detected from inpatients of the ward in the 6 months prior to the outbreak (from October 2018 through March 2019) and in the 6 months following the outbreak (from June through November 2019) was one and two, respec tively.

Characteristics of the patients
All 15 patients with S. pyogenes detection had head and neck cancer (Table 1).The mean age was 68.4 (standard deviation: 10.4) years and eight patients were male.All patients had temporary or permanent tracheostomy.Although all but one patient (Patient-11) was in a four-patient room on the day of detection, only three patients were in the same room within 2 days prior to detection with a patient carrying S. pyogenes before or within 24 hours after the initiation of effective antimicrobial therapy.All patients were detected after more than 7 days of admission, except one patient (Patient-1), who was detected 3 days after admission and was the first patient identified in this outbreak.All patients had a history of procedure room use within 7 days prior to detection of S. pyogenes.There were eight patients with a history of surgery within 30 days, all of whom developed skin and soft tissue infections at the surgical site caused by S. pyogenes.In addition, two patients had skin and soft tissue infections from surgical wounds of earlier operations, one patient had an abscess at the site of an oral tumor, and one patient had pneumonia.All patients with S. pyogenes infection were successfully treated.Of the patients in whom S. pyogenes was detected in culture tests, only three patients were asymptomatic.There were no ICU admissions within 7 days or all-cause deaths within 90 days.

Molecular characteristics of the outbreak isolate
The S. pyogenes isolates from 14 patients and two HCWs were available for wholegenome sequencing analysis.Whole-genome sequencing revealed that all these isolates belonged to emm89 and ST646.SNP analysis was performed on the core genome (95.4% of the 1,741,348 bp genomic sequence of MGAS27061) of these isolates and representa tive isolates of emm89 clade 1, 2, and 3.No SNP differences were found among the isolates detected from patients and HCWs in April, and the two isolates detected from patients in May differed by only a single SNP from those detected in April.The SNP was located in the pyrH gene (locus_tag: MGAS27061_RS01950).The SNP difference between the representative isolate of the outbreak isolates (FUJ00398) and the representative isolates of clade 1, 2, and 3 was 1,868, 374, and 42, respectively, suggesting that the outbreak isolates belong to clade 3. FUJ00398 had the emm89/clade 3-specific promoter sequence of nga-ifs-slo, genes for S. pyogenes NADase (SPN), an endogenous inhibitor of SPN, and streptolysin O (SLO), respectively, and did not carry the hasABC, which is involved in the synthesis of the hyaluronic acid capsule (22).
It was found that genomic sequences of 162 emm89 strains isolated in Japan were available for comparison by referring to the articles identified by a PubMed search performed on 20 February 2024.All but one strain was derived from a single study  (12,23).The initial construction of a phylogenetic tree identified two genetically highly distant strains (OS01 and TK32), a result that was consistent with the previous publication (12).The final phylogenetic tree was created excluding these two strains to allow for the detection of minor genetic differences.The core genome used for phylogenetic analysis was 1,019,443 bp (59.6% of the 1,709,394 bp genomic sequence of the reference strain MGAS23530).Sixty-four ST646 strains, including the representative strain of this outbreak (FUJ00398), diverged from 93 ST101 strains to form a closely related population (Fig. 3).Four of the ST646 descendant strains were further accompanied by a single nucleotide mutation in the gki gene (ST1454).

DISCUSSION
An outbreak of S. pyogenes involving 15 patients and four HCWs occurred in a surgical ward for head and neck cancer patients over a 2-month period, and whole-genome sequencing analysis of the isolates confirmed clonal spread of emm89/clade 3 isolates.Twelve patients developed invasive infections, mainly skin and soft tissue infections at the surgical site, but all were cured without critical illness.All four HCWs presented with mild pharyngitis.
Numerous hospital outbreaks of S. pyogenes have been reported, occurring in ICUs, medical and surgical wards, and obstetric wards (24).Outbreaks are caused by patientto-patient transmission, transmission from an HCW carrying the organism, transmission through a contaminated hospital environment, or a combination of the above.HCWs are reported to be more likely to be involved in transmission in outbreaks in surgical or obstetric wards (24).Although the response to hospital outbreaks of S. pyogenes has not been fully standardized, the UK guidelines suggest the formation of an outbreak control team, epidemiological investigation including retrospective analysis of hospital-onset S. pyogenes infections in the past 6 months, reaffirmation of infection control policies with prompt correction of any flaws, consideration of screening for HCWs and decolonization of positive individuals, identification of common sources of infection (baths, showers, etc.) and subsequent environmental sampling, screening of patients, and chemoprophy laxis (25).In the present case of a ward-wide outbreak, the response was led by the infection control team of the hospital, standard precautions and hand hygiene were reinforced, and screening and empiric antimicrobial administration were performed on all inpatients and HCWs of the ward.In addition, environmental cultures were performed in the procedure rooms, but S. pyogenes was not identified in this outbreak investigation.
Although the outbreak was contained within approximately 2 months, there were two new cases after the initiation of the intervention, suggesting that transmission had persisted via an unidentified source.In this outbreak response, individuals with S. pyogenes isolation who had been treated for at least 24 hours and whose clinical symptoms had improved were not subject to work suspension (HCWs) or isolation in single-patient room and droplet precautions (patients).Although it is generally believed that shedding of S. pyogenes stops within 24 hours of initiation of effective treatment, long-term persistent shedding is possible in practice, and therefore, repeated culture testing after eradication therapy and removal of specific precautions after confirmation of disappearance is suggested in the event of outbreaks (25).In this outbreak, culture testing was not repeated after the initial intervention because there were no new cases for more than 2 weeks after the intervention and the two subsequent cases occurred approximately 2 weeks apart.
Patients with head and neck cancer who have undergone airway modification surgery may be at increased risk of acquiring and spreading S. pyogenes.Frequent airway care from HCWs may increase the risk of transmission from HCWs, and extensive and persis tent droplet spread from the tracheotomy may increase the risk of transmission to surrounding patients and contamination of the hospital environment.In addition, patients with a tracheostomy have difficulty wearing well-fitting masks.In fact, in an outbreak of S. pyogenes emm1 involving three patients with tracheostomy, S. pyogenes was detected from the curtains of the patient rooms and was suspected as a potential source of an outbreak (26).Although environmental disinfection of the patient rooms and the procedure rooms was reinforced in response to this outbreak, the involvement of the hospital environment is unknown because the culture tests were negative.Consider ing the possible presence of unidentified persistent carriers and the difficulty in prevent ing the dispersal of droplets from patients with a tracheostomy, universal masking of HCWs in the head and neck surgery ward, which was introduced in this ward after the COVID-19 pandemic, may be a reasonable option.
The outbreak isolates belonged to emm89/clade 3 and ST646, and the SNP difference was 0-1, confirming clonal spread.S. pyogenes emm89/clade 3 isolates have recently emerged worldwide as one of the major clones of S. pyogenes causing invasive infections (6)(7)(8)(9).emm89/clade 3 isolates emerged in the early 2000s and increased globally in the late 2000s, replacing the previously dominant emm89/clade 1 isolates (27).emm89/ clade 3 isolates differ from emm89/clade 1 and emm89/clade 2 isolates in the nga-ifs-slo promoter sequence and have two SNPs identical to the preceding highly virulent clone, emm1 isolates, leading to increased production of two toxins, SPN and SLO (22).emm89/ clade 3 isolates were also characterized as lacking hasABC.Although it is assumed that loss of the capsule is associated with decreased virulence, it has been reported that increased production of SPN and SLO results in high virulence even without the capsule in a mouse infection model (28).Genomic data analysis of internationally collected S. pyogenes strains revealed that the combination of a promoter sequence with high expression of nga-ifs-slo and a mutation in has gene was also observed in several other common genotypes, including emm28 and emm87, which may suggest that this combination may confer some selection advantage over having only one of these characteristics (29).
To date, emm89/clade 3 has been reported to belong mainly to ST101 worldwide, while the current outbreak isolate was ST646 (6, 7).ST646 is a variant of ST101 with only a single nucleotide mutation of murI among the seven housekeeping gene sequences for MLST.Our phylogenetic analysis for emm89 strains isolated in Japan revealed that ST646 strains diverged from ST101 strains and many clinical strains of this clone have already been isolated.Two of the eight isolates of S. pyogenes emm89 detected in 2009-2013 in patients with balanoposthitis in Nagoya, Japan, were reported to be ST646 (the remaining isolates were ST101) (30).Future studies on the epidemiology and clinical characteristics of the emm89/clade 3 isolate of ST646 in Japan and worldwide are warranted.
This study has several limitations.First, the origin of this outbreak could not be identified.Since S. pyogenes was detected in the first identified patient 3 days after admission to the hospital, it was assumed that the patient developed the disease after an incubation period of community-acquired infection.However, we were unable to confirm the clonality of the isolate from this patient with other isolates because it was not stored, therefore, it is possible that spread among the other patients occurred unrelated to the index patient.Although the timeline of detection suggests that the HCWs acquired S. pyogenes from the patients under their care, we cannot rule out the possibility that there were unidentified asymptomatic individuals earlier in the outbreak who may have been the source of the outbreak.Second, we may have missed some carriers because we did not screen the vagina and anus of the HCWs, and the wound and tracheotomy-aspirated sputum of the patients in the sites of active surveillance culture specimen collection.In addition, we may have missed persistent carriers because we did not confirm the clearance of bacteria with culture after treatment.These interventions were planned to be implemented if the outbreak persisted or if severe cases occurred, but the outbreak was contained without these events.Third, molecular typing of S. pyogenes strains detected before and after the outbreak period was not performed.This may have missed the true onset and termination of the clonal spread of the S. pyogenes strains in the ward.
In conclusion, an outbreak of S. pyogenes emm89/clade 3 ST646 involving postop erative patients and HCWs in a head and neck surgery ward was documented.The outbreak was successfully contained in approximately 2 months after the reinforcement of standard precautions and hand hygiene, and implementation of ward-wide active surveillance culture and empiric antimicrobial administration of patients and HCWs.The global spread of hypervirulent clones may increase the risk of developing infections due to nosocomial transmission of S. pyogenes, and it is necessary to maintain a high level of standard precautions and hand hygiene, especially in wards caring for high-risk populations including patients with head and neck cancer, and to take appropriate early action in case of possible outbreaks.review and editing | Hiroki Mitani, Investigation, Resources, Writing -review and editing | Daisuke Ohkushi, Conceptualization, Supervision, Writing -review and editing

a
Strains were unavailable for one patient (Patient-1) because the strain was not stored and for two nurses (Nurse-1, -2) because they were diagnosed by rapid antigen testing.bAges are shown in 10-year groups.c

FIG 1
FIG 1 Active surveillance cultures and empirical antimicrobial administration performed on (a) hospitalized patients and (b) healthcare workers in response to the outbreak.

FIG 2
FIG 2 Epidemic curve of S. pyogenes detection in the ward.Boxes indicate periods when S. pyogenes was detected.Gray and white boxes represent patients and healthcare workers in whom S. pyogenes was detected, respectively.

FIG 3
FIG3 Phylogenetic tree of emm89 strains isolated in Japan.Allele profile (gki-gtr-murI-mutS-recP-xpt-yqiL) for each ST is indicated in parentheses.The representative strain of this outbreak (FUJ00398) is underlined.ST, sequence type.

TABLE 1
Cases with S. pyogenes detection during the outbreak d