SEARCH

SEARCH BY CITATION

Keywords:

  • Corynebacterium striatum ;
  • epidemiology;
  • MALDI-TOF MS;
  • nosocomial outbreak;
  • typing method

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

During an 8-month period, 24 Corynebacterium striatum isolates recovered from lower respiratory tract specimens of 10 hospitalized patients were characterized. The organisms were identified by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and by 16S rRNA gene sequencing. The cluster of C. striatum exclusively affected patients who had been admitted to an intensive care unit and/or subsequently transferred to one medium-size respiratory care unit. Prolonged duration of hospitalization, advanced stage of chronic obstructive pulmonary disease, recent administration of antibiotics and exposure to an invasive diagnostic procedure were the most commonly found risk factors in these patients. Seven patients were colonized and three infected. All strains displayed a similar broad spectrum resistance to antimicrobial agents, remaining susceptible to vancomycin only. Typing analysis by MALDI-TOF MS and by semi-automated repetitive sequence-based PCR (DiversiLab typing) showed that all outbreak-associated C. striatum isolates clustered together in one single type while they differed markedly from epidemiologically unrelated C. striatum isolates. Pulsed-field gel electrophoresis (PFGE) profiles revealed three distinct PFGE types among the C. striatum isolates associated with the outbreak while all external strains except one belonged to a distinct type. We conclude that C. striatum is an opportunistic nosocomial pathogen in long-term hospitalized patients and can be at the origin of major outbreaks. The routine use of MALDI-TOF MS greatly facilitated the recognition/identification of this organism in clinical samples and this technique could also offer the potential to be used as an easy and rapid epidemiological typing tool for outbreak investigation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

The genus Corynebacterium refers to gram-positive bacteria that form part of the normal microbiota of human skin and mucous membranes and that are also widely distributed in the environment [1, 2]. The significance and prevalence of coryneform bacteria are not always well established.

Corynebacterium striatum has been isolated as part of the normal human nasopharyngeal flora and from the skin. However, this organism has also been shown to be potentially pathogenic in patients with chronic diseases and in specific circumstances (e.g. following repeated exposure to broad-spectrum antibiotics, after the use of invasive medical procedures and in the presence of organic obstructive pathologies) [3-5]. Most reported C. striatum infections involved the respiratory tract and several outbreaks of nosocomial infections have been described [6-9].

At the Mont-Godinne University Hospital (Yvoir, Belgium) we recovered 24 C. striatum strains from respiratory samples of 10 patients hospitalized in the intensive care unit (ICU) and at one medical respiratory unit over an 8-month period.

To understand the possible outbreak sources and transmission routes we reviewed the medical records of all patients from whom C. striatum was isolated over the study period. Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) was used for strain identification and was challenged as an epidemiological typing tool through the interpretative reading of the strain spectra in the form of a generated dendrogram. Results were compared with semi-automated repetitive-sequence-based PCR (rep-PCR) (DiversiLab, bioMérieux, Marcy-l'Etoile, France) and PFGE (pulsed-field gel electrophoresis).

Material and Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

Patients

We reviewed the medical records of all patients from whom C. striatum isolates had been recovered from clinical specimens for the presence of putative risk factors for nosocomial acquisition. The study took place at UCL Mont-Godinne University Hospital (Yvoir, Belgium), a 430-bed teaching hospital with a laboratory mostly receiving samples from the hospital. The status of colonization or infection was assessed according to the CDC criteria for definition of infection [10].

Bacterial isolates

From January 2011 to August 2011, all routine specimens yielding the isolation of C. striatum at culture and considered to be of potential clinical significance (i.e. isolation from biologically sterile fluids or as a pure or predominant culture growth from non-sterile clinical specimens) were collected but only the first strain for each patient was selected for the different analyses.

In parallel, five epidemiologically unrelated C. striatum clinical isolates, recovered from respiratory samples, collected from other Belgian hospitals between 2000 and 2010 and sent for confirmation of identification to a reference laboratory (Professor G. Wauters, Microbiology Unit, Faculty of Medecine, Catholic University of Louvain, Brussels, Belgium) were also included in the analysis for the purpose of comparison.

All C. striatum isolates were initially identified through the clinical laboratory routine workflow by MALDI-TOF MS and their identification was further confirmed by sequencing of the entire 16S rRNA gene according to a previously published method [11].

In vitro susceptibility to ten antimicrobial agents (penicillin, ampicillin, erythromycin, clindamycin, gentamicin, tetracycline, ciprofloxacin, co-trimoxazole, rifampicin and vancomycin) was assessed by broth microdilution according to Clinical and Laboratory Standards Institute guidelines [12].

MALDI-TOF MS

MALDI-TOF MS measurements were realized on a microflex LT (Bruker Daltonik, Bremen, Germany). Spectra were recorded by the positive linear method in a mass range from 2000 to 20000 Da.

Strain identification

Single colonies were spotted on a steel target overlaid with 1 μl matrix solution dissolved in a basic organic solvent composed of 50% acetonitrile and 2.5% tri-fluor-acetic-acid. The acquired bacterial spectra with MALDI-TOF MS were analysed in the MALDI Biotyper 2.0 software database, leading to scored identification results. According to the specifications of the manufacturer, a high log score ≥2 was required for identification to species level.

Dendrogram creation

Approximately 10 colonies of each C. striatum isolate were scraped from a 24-h culture on trypticase soy blood agar and added to 500 μL distilled water followed by an ethanol-formic acid extraction procedure. One μl of the final extraction product was spotted eight times onto a steel target and was overlaid with 1 μl of matrix solution. Each spot was measured three times with the MBT_FC.par flexControl method and the MBT-autoX.axe autoExecute method. The resulting 24 spectra were downloaded into the MALDI BioTyper software and assembled in order to generate a single mean spectrum accounting for the extracted strain. In order to appreciate the correlation between the organisms and visualize the clustering, a dendrogram was calculated.

DiversiLab typing

DNA extraction for all C. striatum strains was achieved with the UltraClean microbial DNA isolation kit (MoBio, Carlsbad, CA, USA) followed by DNA quantification with a NanoDrop 2000/2000C spectrophotometer (Thermofisher, Friendship, ME, USA). The DiversiLab Bacterial fingerprint kit (Ref# 270633) was used according to the manufacturer's instructions to perform the rep-PCR amplification. PCR products were detected and sized on an Agilent 2100 bioanalyser (Agilent Technologies, Diegem, Belgium). Sample clustering was studied by comparing the strain patterns looking for band differences and the percentage similarity analysed. Isolates with identical strain patterns were considered indistinguishable if the similarity percentage was ≥97% [13].

PFGE

For PFGE analysis, C. striatum DNA was extracted using lysozyme (20 mg/mL) and achromopeptidase (500 Units/mL). Macrorestriction (XbaI) patterns were analysed using Dice coefficient with Bionumerics software (Applied Maths, Kortrijk, Belgium). The classification criteria for PFGE analysis included type, designated by a capital letter and patterns showing 0-3 DNA fragment difference corresponding to a similarity cut-off of ≥80% [14].

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

Bacterial isolates and medical records

From January 2011 to August 2011, we recovered among all routine culture microbiological samples a total of 24 clinical isolates of C. striatum. All strains originated from lower respiratory tract specimens of 10 distinct hospitalized patients. The 24 strains were correctly identified to the species [log(score) ≥2] with MALDI-TOF MS. 16S rRNA gene sequences of the 10 analysed strains (one strain/patient) showed ≥99.6% similarity with the sequence from C. striatum type strain NCTC 764T. All C. striatum clinical isolates displayed an identical multidrug resistance pattern with susceptibility to vancomycin only. The five epidemiologically unrelated C. striatum strains showed a less resistant profile as all isolates were susceptible to penicillin, ampicillin, erythromycin, clindamycin and vancomycin while showing heterogenous profiles to the other antibiotics (data not shown).

The clinical characteristics of the patients are presented in Table 1. According to CDC criteria, seven patients were considered to be colonized and three infected. All affected patients had been hospitalized for prolonged periods before the first isolation of C. striatum (mean, 33 days; range, 5–90 days).

Table 1. Summary of the medical records of the 10 patients with positive C. striatum isolates in respiratory samples
PatientAge/genderUnderlying conditionUnitPrevious antibiotic(s)Fibroscopy IntubationC. striatum statusAssociated strainMedical devices Antibiotic treatmentIssue
  1. AMC, amoxicillin/clavulanate; AMI, amikacin; AZT, azithromycin; CAZ, ceftazidime; CIP, ciprofloxacin; COPD, chronic obstructive pulmonary disease; CVVH, continuous veno-venous haemofiltration; CXM, cefuroxime-axetil; FEP, cefepim; IV, intravenous; LVX, levofloxacin; MIN, minocycline; MRP, meropenem; PTZ, piperacillin + tazobactam; SL, colistine; SXT, co-trimoxazole; F, female; M, male. P. aeruginosa, Pseudomonas aeruginosa; S. maltophilia, Stenotropho monas maltophilia.

162/MLobectomy Lung, carcinomaIntensive care YesYesPneumoniae, Empyema Chest drainPTZ CXMRecovered
259/FLung transplant, COPDMedical unitAZTYesYesColonization Gastric catheterRecovered
391/FCorticotherapy, Follicular lymphomaMedical unitPTZNoNoColonization Recovered
479/MCOPDMedical unitMRP+AMIYesNoColonization IV catheterRecovered
588/MCOPDMedical unit

AMC

LVX

YesNoPneumoniaeP. aeruginosaCAZRecovered
665/MLung transplant, COPDIntensive care

PTZ

CAZ

YesYesPneumoniae

P. aeruginosa

S. maltophilia

IV, arterial and dialysis catheter

Tracheostomy

CIP, MIN

CS

Recovered
761/MSurgery for aortic dissectionIntensive care

CIP

AMC

YesYesColonization

Aortic prosthesis

Chest drain

Recovered
873/FMitral valve replacementIntensive care

AMC

FEP

YesYesColonizationP. aeruginosa

CVVH, IV and arterial catheter

Artificial mitral valva

Died
964/MDiabetes, COPDMedical unit NoNoColonization Oesophageal prosthesisRecovered
1075/FCOPDIntensive careCIP, SXTYesNoColonization

Chest drain

Naso-gastric tube

Recovered

During the outbreak, at least one colonized or infected patient was hospitalized in the ICU or in the medical respiratory unit at the time by which a new carrier of C. striatum was detected (Fig. 1). Eight of the 10 patients had major co-morbidities and had been exposed to multiple previous courses of antibiotic treatment. Eight patients did undergo a bronchial fibroscopy and five patients were intubated. The medical care of eight patients required the use of invasive medical devices. Finally, nine patients were discharged from the hospital while one colonized patient died during his stay.

image

Figure 1. Hospital route of the 10 C. striatum colonized/infected patients between January and August 2011 □, presence in intensive care unit; ○, presence in medical respiratory unit; grey symbols, respiratory samples that were culture-negative for C. striatum; black symbols, respiratory samples that were culture-positive for C. striatum.

Download figure to PowerPoint

Epidemiological analysis

Dendrogram with MALDI-TOF MS

The dendrogram (Fig. 2) generated from the 15 registered C. striatum spectra pointed out a distinctive cluster assembling the 10 outbreak-related C. striatum strains and the external strains 4 and 5 with a maximum distance level of 100. The acquired spectra of three out of the five unrelated tested C. striatum external strains showed a distinct spectrum pattern.

image

Figure 2. Dendrogram of the C. striatum strains of the 10 colonized/infected hospital patients and of five external C. striatum strains. Cs, Corynebacterium striatum.

Download figure to PowerPoint

DiversiLab typing

All 10 outbreak-related C. striatum isolates had the same repPCR patterns and had ≥97.7% similarity level between each other while the five epidemiologically non-related strains displayed different strain patterns, with similarity levels <97% between each other (Fig. 3).

image

Figure 3. DiversiLab typing and PFGE clone types of the C. striatum strains of the 10 colonized/infected hospitalized patients and five epidemiologically unrelated strains (‘external’ C. striatum strains). Cs, Corynebacterium striatum; DL, DiversiLab; PFGE, pulsed-field gel electrophoresis.

Download figure to PowerPoint

PFGE

PFGE delineated the 10 C. striatum patient strains into three distinct PFGE types (Fig. 3). Type A comprised the outbreak-related C. striatum isolates from patients 1 to 7 and C. striatum external isolate 5. Type F comprised the C. striatum isolates from patients 8 and 9. Finally Type H was a singleton only including the C. striatum patient 10 strain.

The analysis of the four leftover C. striatum epidemiologically non related strains led to four singleton PFGE types (B, C, D and G).

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

Over recent years Corynebacteria have been recognized as opportunistic pathogens able to cause various types of healthcare-associated infections in immunocompromised hosts [2]. It is therefore essential for each clinical microbiology laboratory to use techniques that identify correctly and rapidly these bacteria. Identification of Corynebacteria by conventional methods is suboptimal and it is likely that their true prevalence in clinical specimens either as colonizers or as pathogens is largely underestimated. MALDI-TOF MS has proven accuracy for rapid identification of toxigenic Corynebacterium species and non-diphtheria Corynebacterium, including C. striatum [15-17].

In our setting, the repeated identification of C. striatum in clinical respiratory samples of several hospitalized patients over a limited period of time raised the suspicion of an ongoing outbreak. While only one C. striatum isolate had been recovered from a hospitalized patient in the preceding 8-month baseline period, we identified 24 clinical C. striatum isolates in 10 distinct hospitalized patients over the 8 following months, corresponding to an attack rate of 10 to 1. Following this observation, we established the ‘outbreak case definition’ as each hospitalized patient with a clinical sample growing mainly or exclusively C. striatum colonies. A retrospective investigation of the patients' medical records was carried out to describe the outbreak population. Several risk factors were identified in our patients. The most frequently found underlying co-morbidity in our series was chronic obstructive pulmonary disease for eight of the ten patients. It is very likely that such long-lasting chronic disease along with the frequent use of respiratory equipment did contribute to the respiratory tract colonization and the infectious potential of this organism in immunocompromised patients. Previous exposure to antibiotic treatment observed in eight patients also most likely contributed to the overgrowth of this multidrug-resistant organism through selective pressure.

Upon the case definition and the subsequent analysis of risk factors, an epidemic curve (date not shown) and a spatial/temporal distribution (Fig. 1) were drawn, aiming to understand the routes of transmission and sources of the C. striatum outbreak. Figure 1 highlights the particularity of this outbreak, which was ongoing simultaneously in two units despite the fact that the medical and ICU wards were geographically separated and that each unit was staffed with distinct medical and nursing teams. The colonized/infected patients probably constituted the reservoirs of transmission because all had a very long hospital stay and at least one colonized patient was always present in one of the two affected units during the outbreak period at the time by which a new case of C. striatum was recognized. We considered as the most probable scenario that patient #2 who had been nursed in contact with the index case (patient #1) in the ICU in January 2011 and who was subsequently transferred to the respiratory medical unit, was at the origin of the dissemination of this strain between the two units. The healthcare staff in both units are strongly suspected to be the vector of cross-transmission of the C. striatum strain between patients. In a previous report, Brandenburg et al. reported the 12-month persistence of a single C. striatum epidemic strain in patients in a surgical ICU, suggesting that asymptomatically colonized patients may constitute the main reservoir of C. striatum and that patient-to-patient transmission probably occurs via the hands of the personnel [7]. Our study hereby highlights the protracted and silencious pattern of a C. striatum outbreak. Based on all these elements, an alert was issued by the infection control team to reinforce hand hygiene practice and barrier precautions.

In parallel, a laboratory investigation of the outbreak-related C. striatum strains was performed. Several molecular typing methods (DNA fingerprinting and genotyping) have been used to demonstrate an epidemiological link between the isolates in previously described C. striatum outbreaks [6-8]. These methods nevertheless require highly qualified staff and expensive dedicated material and they usually require several days or weeks before results are obtained and a feedback is given to the units. MALDI-TOF MS has proven to be reliable for identification of bacteria but as yet there have been only few studies that have assessed the value of this technique as an epidemiological typing tool in the setting of a noscomial outbreak [18]. In our study we aimed to set up an easy and fast subtyping technique accessible to any MALDI-TOF MS user. The technique was based on an ethanol-formic acid strain extraction method (routine test for difficult strains) and a dendrogram cluster analysis only requiring the Maldi BioTyper Software without any additional statistical tool. We determined the diversity of the C. striatum isolates collected in the setting of our local outbreak and compared these strains with epidemiologically unrelated isolates collected from specimens in several hospitals. All outbreak-related strains clustered in a single clone both by MALDI-TOF MS dendrogram and with the DiversiLab typing method. Yet, two epidemiologically unrelated isolates (external C. striatum strains #4 and #5 also) clustered within the outbreak cluster in the dendrogram while they corresponded to a different clone type with DiversiLab. PFGE analysis subdivided the ten outbreak-associated C. striatum strains into three clusters (A, F and H). Cluster A included the first seven patient strains and C. striatum external strain #5, an observation also made with the MALDI-TOF MS dendrogram. Interestingly, by PFGE the strains were found to be grouped into three different types according to chronological sequence (time) of isolation (i.e. cluster A included patient strains #1 to #7, cluster F strains #8 and #9, and finally cluster H strain #10). As the duration of this outbreak had spanned over an 8-month period, it may well be possible that the C. striatum outbreak strain had undergone some chromosomal changes that might have been detected by PFGE analysis, considered as a highly sensitive molecular epidemiological tool. In conclusion, typing results by MALDI TOF MS were subsequently confirmed by intergenic rep PCR typing analysis but both methods appeared slightly less discriminant than PFGE for recognition and differentiation of certain clonal lineages. Based on a diversity analysis of C. striatum strains by MALDI-TOF MS and by a multigenic sequencing approach, Gomila et al. similarly suggested that MALDI-TOF MS could be an efficient tool for discrimination of bacterial strains below the species level though not as discriminatory as the other molecular typing techniques [17].

A hospital outbreak is never a coincidence but a reflection of a failure in the infection control measures. In line with several previous reports, our study once more highlights the role of C. striatum in nosocomial outbreaks mostly affecting immunocompromised patients [4-9].

Based on our experience, we believe that rapid confirmation of the spread of a single clone/type of a nosocomial pathogen within or across several hospital units by a locally available typing method such as MALDI-TOF MS may contribute to improving the management of an outbreak. This study is one of the first to document the potential usefulness of MALDI TOF MS as a subtyping tool for investigating nosocomial outbreaks. Further studies are, however, needed to confirm our observations.

Transparency Declaration

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References

The authors declares no conflicts of interest.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Material and Methods
  5. Results
  6. Discussion
  7. Transparency Declaration
  8. References
  • 1
    Sommerville DA. A taxonomic scheme for aerobic diphtheroids from human skin. J Med Microbiol 1973; 6: 215223.
  • 2
    Bernard K. The genus Corynebacterium and other medically-relevant Coryneform like bacteria. J Clin Microbiol 2012 Oct; 50: 31528.
  • 3
    Martinez-Martinez L, Suarez AI, Rodriguez-Banõ J, Bernard K, Muniain MA. Clinical significance of Corynebacterium striatum isolated from human samples. Clin Microbiol Infect 1997; 6: 634639.
  • 4
    Lee PP, Ferguson DA Jr, Sarubbi FA. Corynebacterium striatum: an underappreciated community and nosocomial pathogen. J Infect 2005; 50: 338343.
  • 5
    Wong KY, Chan YC, Wong CY. Corynebacterium striatum as an emerging pathogen. J Hosp Infect 2010; 76: 354372.
  • 6
    Leonard RB, Nowowiejski DJ, Warren JJ, Finn DJ, Coyle MB. Molecular evidence of person-to-person transmission of a pigmented strain of Corynebacterium striatum in intensive care units. J Clin Microbiol 1994; 1: 164169.
  • 7
    Brandenburg AH, van Belkum A, van Pelt C, Bruining HA, Mouton JW, Verbrugh HA. Patient-to-patient spread of a single strain of Corynebacterium striatum causing infections in a surgical intensive care unit. J Clin Microbiol 1996; 9: 20892094.
  • 8
    Otsuka Y, Ohkusu K, Kawamura Y, Baba S, Ezaki T, Kimura S. Emergence of multidrug-resistant Corynebacterium striatum as a nosocomial pathogen in long-term hospitalized patients with underlying diseases. Diagn Microbial Infect Dis 2005; 54: 109114.
  • 9
    Renom F, Garau M, Rubi M, Ramis F, Galmes A, Soriano JB. Nosocomial outbreak of Corynebacterium striatum infection in patients with chronic obstructive pulmonary disease. J Clin Microbiol 2007; 6: 20642067.
  • 10
    Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of healthcare associated infection and criteria for specific types of infection in the acute care setting. Am J Infect Control 2008; 36: 309332.
  • 11
    Wauters G, Avesani A, Laffineur K et al. Brevibacterium lutescens sp. nov., from human and environmental samples. Int J Syst Evol Microbiol 2003; 53: 13211325.
  • 12
    Clinical and Laboratory Standards Institute. Methods for antimicrobial dilution and disk susceptibility testing of infrequently isolated or fastidious bacteria. CLSI document M45-2A. Wayne, PA: CLSI, 2010.
  • 13
    Deplano A, Denis O, Rodriguez-Villalobos H, De Ryck R, Struelens MJ, Hallin M. Controlled performance evaluation of the diversiLab repetitive-sequence-based genotyping system for typing multidrug-resistant health care associated bacterial pathogens. J Clin Microbiol 2011; 49: 36163620.
  • 14
    Denis O, Deplano A, De Ryck R, Nonhoff C, Struelens MJ. Emergence and spread of gentamicin-susceptible strains of methicillin-resistant Staphylococcus aureus in Belgian hospitals. Microbial Drug resistance 2003; 9: 6171.
  • 15
    Konrad R, Berger A, Huber I et al. Matrix-assisted laser desorption/ionisation time-of-flight (MALDI-TOF) mass spectrometry as a tool for rapid diagnosis of potenitally toxigenic Corynebacterium species in the laboratory management of diphtheria-associated bacteria. Euro Surveill 2010; 15: 19699.
  • 16
    Alatoom AA, Cazanave JC, Cunningham SA, Ihde SM, Patel R. Identification of non-diphtheriae Corynebacterium by use of matrix-assisted laser desorption ionisation – time of flight mass spectrometry. J Clin Microbiol 2012; 1: 160163.
  • 17
    Gomila M, Renom F. Gallegos MdC et al. Identification and diversity of multi-resistant Corynebacterium striatum clinical isolates by MALDI-TOF mass spectrometry and by mutigene sequencing approach. BMC Microbiol 2012; 12: 5260.
  • 18
    Griffin MP, Gareth RP, Schooneveldt MJ et al. Use of Matrix-assisted laser desorption ionization-time of flight mass spectrometry to identify vancomycin-resistant Enterococci and investigate the epidemiology of an outbreak. J Clin Microbiol 2012; 9: 29182931.