Evaluation of VITEK 2 compact and VITEK MS in the identification of coagulase-negative staphylococci isolated from blood cultures

1. Ghanem Laboratório Clínico, Santa Catarina, Brazil
2. Laboratório Médico Santa Luzia, Santa Catarina, Brazil
3. Universidade Regional de Blumenau, Santa Catarina, Brazil

Corresponding author

Alessandro Conrado de Oliveira Silveira
Universidade Regional de Blumenau
Rua São Paulo, 2.171
CEP: 89030-001; Blumenau-SC, Brasil
e-mail: alessandrocosilveira@gmail.com

INTRODUCTION

In 2014, 47 species and 23 subspecies of the Staphylococcus genus were described, with the S. aureus being the best known, and frequently associated with a series of infections and severe intoxications in human beings^{(1, 2)}. Most species belong to the group of coagulase-negative staphylococci (CNS), which for many years were considered skin-colonizing or rarely pathogenic bacteria⁽²⁾.

CNS are among the most common microorganisms isolated from blood stream infections. They are, in general, connected with foreign body infections, being classified as healthcareassociated infections⁽³⁾. However, they are some of the most common blood culture contaminants, thus the use of different clinical and laboratory criteria is necessary to distinguish between contamination and clinically significant bacteremia⁽⁴⁾.

In most clinical laboratories, CNS are classified through their biochemical characteristics using traditional and/or automated phenotypic methods. Nevertheless, identification by a phenotypic method is a great challenge, despite the improvement in identification by automated systems in the latest years. Besides taxonomy of the Staphylococcus genus has been evolving over the years, the use of commercial kits based on biochemical reactions does not produce reliable results for CNS identification in clinical samples^(5-7). Several automated methods based on phenotypic characteristics have been developed to improve identification, but there are limitations because they depend on the expression of metabolic activities and/or morphologic characteristics, and, principally, due to phenotypic differences between isolates of the same species^{(1, 2, 8)}.

There have been new technologies to devise more rapid precise and sensitive methods aimed at enhancing assistance to patient and therapy, as, for instance, genetic sequencing, which has a highly discriminatory power of diagnosis, uncovering numerous errors in phenotypic identification and becoming, as a result, the gold standard method for microorganism identification⁽⁹⁾. However, such methods are employed mainly by reference laboratories and require specialized instrumentation, dedicated laboratory space and highly trained staff. Such restrictions are unfeasible for most laboratories, and because of that, the use of automated systems based on phenotypic analysis still predominates, despite imperfections in precision, robustness and the need for confirmatory testing⁽⁹⁾.

The Matrix Assisted Laser Desorption/Ionization-Time of Flight (MALDI-TOF) method comes as an alternative for biochemical analysis and identification based on gene sequencing⁽¹⁰⁾. The main advantage of MALDI-TOF technology over other laboratory techniques for microorganism identification is that it is rapid to obtain results. From inoculum preparation to final reading, an isolated result can be obtained in less than 30 minutes⁽¹¹⁾. Nowadays two instruments like these are available in the market and registered at the Ministry of Health: Microflex LT™ (Bruker Daltonics/BD Diagnostics) and VITEK MS™ (bioMérieux)⁽¹²⁾.

Therefore, this study aims principally to evaluate the identification of CNS isolated in the laboratory routine by means of an automated phenotypic method and mass spectrometry, so as to compare both methods.

MATERIALS AND METHOD

The microorganisms came from whole blood samples collected between September 2015 and May 2016. Because this is a comparative study, there were not clinical criteria for sample selection. Samples of a same patient were rejected. All microorganisms that had an index greater than or equal to 90% confidence level were included in the study. In the VITEK^® 2 Compact device, identification was made by the Advanced Colorimetry™ technology, while VITEK MS used MALDI-TOF technology.

Blood culture vials were incubated in the BacT/ALERT^® (bioMérieux) instrument for a maximum period of five days. Positive samples in this period were streaked onto growth media sheep-blood agar and chocolate agar (bioMérieux). The plates were incubated in bacteriological incubator at 35ºC ± 1ºC for 18-24 hours.

The identification of all CNS isolated in the procedure was firstly made by VITEK 2 Compact according to the manufacturer’s manual. GP TEST KIT VTK 2 card was used for identification of all Gram-positive cocci.

The isolates already identified were stored in 20% Skim Milk growth medium and frozen at -20ºC for later identification by VITEK MS. Following, the stored samples were thawed, activated in brain heart infusion (BHI) broth, and streaked in sheep-blood agar (bioMérieux). Microorganisms were then identified by VITEK MS. Later, the results yielded by identification with the automated phenotypic method were compared to those from MALDI-TOF MS, so as to compare both methods.

RESULTS

A total of non-repeated 133 isolates from blood cultures were analyzed. CNS identification was made by an automated phenotypic method and proteomic analysis with VITEK 2 Compact and VITEK MS devices, respectively.

Both instruments had a 92.8% (117/126) agreement, of which 56 were Staphylococcus epidermidis, 34 Staphylococcus hominis ssp. hominis, 15 Staphylococcus haemolyticus, 10 Staphylococcus capitis, one Staphylococcus cohnii ssp. urealyticus, and one Staphylococcus sciuri. All the 133 isolates were identified by VITEK^® 2 Compact as CNS, while VITEK MS did not identify seven samples; there was uncertainty between three species in an isolate.

The isolates not identified by MALDI-TOF technology were reassessed in duplicate, and still their characterization was not possible. There was an inconclusive result in the identification of a microorganism by MALDI-TOF, with uncertainty between Staphylococcus xylosus, Staphylococcus saprophyticus and Staphylococcus cohnii ssp. urealyticus. This same microorganism was identified by another method as Staphylococcus xylosus, with 95% confidence level.

The isolates with the wider disagreement in identification were Staphylococcus hominis ssp. hominis and Staphylococcus epidermidis. The microorganism Staphylococcus lugdunensis, identified by the phenotypic method, was identified by MALDITOF as Staphylococcus aureus, this being a coagulase-positive staphylococcus. All conflicting results were found at the species level (Table).

DISCUSSION

In the latest years, CNS identification was commonly carried out by phenotypic methods through automated commercial systems that many times could not differentiate the variable expression of some staphylococcal phenotypic characteristics. The S. epidermidis group is characterized by the pronounced genomic diversity, while other species such as S. haemolyticus, S. lugdunensis and S. schleiferi present less diversity as pulsed-field gel electrophoresis shows⁽¹⁾.

Staphylococcus epidermidis is the most common species found in infections, followed by S. hominis, S. haemolyticus and S. capitis⁽¹⁾. The group S. epidermidis comprises a common pathogen, needing a rapid decision each time it is found in clinical samples, because it can represent a true infection or just colonization.

Layer et al. (2006)⁽²⁾ conducted a study assessing CNS identification with different methods, which showed 95.24% agreement with the gold standard method (VITEK 2 Compact, bioMérieux) and 76.19% (Phoenix^®, BD Diagnostics, Sparks) for S. epidermidis. Another study, conducted by Ligozzi et al. (2002) ⁽¹³⁾, demonstrated that 86 out of the 100 CNS isolates were identified at the species level, of which S. epidermidis was identified with 92.7% of accuracy.

Spanu et al. (2003)⁽¹⁴⁾ demonstrated that VITEK^® 2 Compact identified correctly 90.5% of CNS and clearly indicated there is still necessity of improvement in the identification of some species by the phenotypic method, especially in S. hominis. isolates. This fact can explain discrepancies observed in our study, in which S. hominis appears in seven out of the nine disagreeing isolates.

MALDI-TOF was between three species in an isolate identified by VITEK^® 2 Compact as Staphylococcus xylosus. The identified species were S. xylosus, S. saprophyticus and S. cohnii ssp. urealyticus.

A study carried out by Delport et al. (2015)⁽¹⁵⁾ compared the CNS identification with MALDI-TOF, VITEK^® 2 Compact and gas chromatography and used strains previously identified by the Kloos and Schleifer^{(16, 17)} classification scheme. A Staphylococcus saprophyticus was identified as Staphylococcus xylosus by MALDITOF and gas chromatography; it was confirmed by rRNA16S sequencing as Staphylococcus xylosus. Another study, performed by Lee et al. (2013)⁽⁷⁾, compared identification of CNS from urine isolates among several methods and verified that two isolates of Staphylococcus cohnii were identified as Staphylococcus xylosus by VITEK^® 2 Compact with 95% confidence level.

In general, novobiocin-resistant staphylococci are reported as Staphylococcus saprophyticus, but nowadays other CNS species that are also novobiocin-resistant have been recognized, such as S. saprophyticus, S. cohnii subsp. cohnii, S. cohnii subsp. urealyticus, S. sciuri and S. xylosus^{(18, 19)}. According to McTaggart (1989)⁽²⁰⁾, S. saprophyticus presumptively identified just by Gram stain, catalase, coagulase, novobiocin resistance and DNAse reaction had biochemical profiles similar to species S. cohnii, S. chrornogenes, S. intermedius, S. xylosus and S. simulans, which share phenotypic similarities with S. saprophyticus. This finding illustrates heterogeneity of CNS resistant to novobiocin. Such resistance is generally encoded by plasmids and thus, inter-species transfer of novobiocin-resistant genes is possible.

An isolate of Staphylococcus lugdunensis identified by VITEK^® 2 Compact with 95% confidence level was identified by VITEK MS as Staphylococcus aureus. Identification of S. lugdunensis by phenotypic method may reveal the presence of specific surface proteins, leading to positive slide coagulase and clumping factor tests, suggesting S. aureus. Although classified as CNS, S. lugdunensis has pathogenic features similar to those of S. aureus, therefore it requires the same antibiogram interpretative criteria as those recommended for S. aureus and different from other CNS⁽²¹⁾. For this reason, the correct species discrimination is important on laboratory routine for characterization of susceptibility testing.

Studies show the efficiency of mass spectrometry technology, relating it to phenotypic methods with the same or better accuracy in identification. Dupont et al. (2010)⁽²²⁾ compared CNS identification between systems VITEK^® 2 and Phoenix^® (Becton, Dickinson) – 234 CNS (20 species) were isolated -. The study used as an identification reference the sodA gene sequencing. On the whole, VITEK MS, VITEK^® 2 and Phoenix^® correctly identified 93.2%, 75.6% and 75.2% of the isolates, respectively. Another study, now using BioTyper 2.0 software instead of the in-house database, compared the results of VITEK MS with the sequencing of the tuf gene for identification of 62 CNS. All the isolates in this study were identified at species level, demonstrating a 100% agreement with the reference method for CNS identification⁽²³⁾. Therefore, VITEK MS showed high precision in identification of CNS at species level, in a simple and rapid form.

In the study performed by Loonen et al.(2012)⁽²⁴⁾,five methods were compared to evaluate accuracy in CNS identification. Results showed identification of 85.9%, 92.3%, 70.4%, 93% and 99.3% for ID 32 Staph strip (bioMérieux), VITEK 2 (GP TEST KIT VTK 2), partial sequencing of RNA 16S (MicroSeq; Applied Biosystems), partial sequencing of the tuf gene (in-house) and Microflex LT MS (Bruker Daltonics), respectively. MALDI-TOF presented superiority in CNS identification in this study, in which 99.3% of isolates were correctly identified.

All CNS, except Staphylococcus lugdunensis, have the same criteria for susceptibility testing recommended by the document M100-S26 of Clinical Laboratory Standards Institute (CLSI) – Performance Standards for Antimicrobial Susceptibility Testing -. In these cases, incorrect identification of CNS affects mainly the emission of epidemiological data to the hospital infection control service and the clinical interpretation in case of two or more different CNS in blood cultures, showing a collection-related contamination. The erroneous identification of Staphylococcus lugdunensis as another CNS can result in therapeutic failure, because the criteria of the antibiogram cut-off point are the same as those of S. aureus and are rarely considered contaminants in sterile sites.

In this study there was 92.8% agreement in CNS identification between both proposed methods. The non-identification of the seven isolates by VITEK MS can be associated with: 1) an interferent in the slide preparation; 2) contamination in the isolate streaking; or 3) incubation time.

As a limitation of this study, we can refer the fact that it does not have a gold standard method to evaluate identifications between both instruments, as well as to analyze accuracy of each method.

CONCLUSION

After comparison between the automated phenotypic method and proteomic analysis, based on studies comparing both methods with a gold standard method, a phenotypic similarity was observed between isolates of the same species, this being the main limitation of the phenotypic method. The technology of mass spectrometry displayed high discrimination in species identification.

Differentiation between CNS species becomes more crucial in blood cultures, in which the presence of two CNS in different samples can rule out a true bacteremia.

With descriptions of many new species found in the last two decades and improved approaches to identification in diagnosis, other microorganisms of the group became evident. At an era in which proteomics and the whole genome sequencing have been established, such species can be increasingly better identified.

REFERENCES

1. Becker K, Heilmann C, Peters G. Coagulase-negative staphylococci. Clin Microbiol Rev. 2014; 24(4): 870-926.

2. Layer F, Ghebremedhin B, Moder KA, König W, König B. Comparative study using various methods for identification of Staphylococcus species in clinical specimens. J Clin Microbiol. 2006; 44(8): 2824-30.

3. Delmas J, Chacornac JP, Robin F, Giammarinaro P, Talon R, Bonnet R. Evaluation of the Vitek 2 system with a variety of Staphylococcus species. J Clin Microbiol. 2008; 46(1): 311-3.

4. Rigatti F, Tizotti MK, Hörner R, et al. Bacteremias por Staphylococcus coagulase negativos oxacilina resistentes em um hospital escola na cidade de Santa Maria, Estado do Rio Grande do Sul. Rev Soc Bras Med Trop. 2010; 43(6): 686-90.

5. Carretto E, Barbarini D, Couto I, et al. Identification of coagulasenegative staphylococci other than Staphylococcus epidermidis by automated ribotyping. Clin Microbiol Infect March. 2005; 11(3): 177-84.

6. Paim TGS, Reiter KC, Oliveira CF, D’Azevedo PA. MALDI-TOF MS performance to identify gram-positive cocci clinical isolates in Porto Alegre/RS, Brazil. J Infect Control. 2013; 2(2): 112-6.

7. Lee TF, Lee H, Chen CM, et al. Comparison of the accuracy of matrix-assisted laser desorption ionization-time of flight mass spectrometry with that of other commercial identification systems for identifying Staphylococcus saprophyticus in urine. J Clin Microbiol. 2013; 51(5): 1563-6.

8. Dubois D, Leyssene D, Chacornac JP, et al. Identification of a variety of staphylococcus species by matrix-assisted laser desorption ionization-time of flight mass spectrometry. J Clin Microbiol. 2010; 48(3): 941-5.

9. Clark AE, Kaleta EJ, Arora A, et al. Matrix-assisted laser desorption ionization-time of flight mass spectrometry: a fundamental shift in the routine practice of clinical microbiology. Clin Microbiol Rev. 2013; 26(3): 547-603.

10. Stevenson LG, Drake SK, Murray PR. Rapid identification of bacteria in positive blood culture broths by matrix-assisted laser desorption ionization-time of flight mass spectrometry. J Clin Microbiol. 2010; 48(2): 444-7.

11. Croxatto A, Prod’hom G, Greub G. Applications of MALDI-TOF mass spectrometry in clinical diagnostic microbiology. FEMS Microbiol Rev. 2012; 36(2): 380-407.

12. Almeida JN, Gioia TSR, Doi AM, Rossi F. Aplicação da tecnologia de espectrometria de massa MALDI-TOF em laboratórios de microbiologia clínica. Microbiologia in foco. 2014. Informativo sbm ano 5/2014.

13. Ligozzi M, Bernini C, Bonora MG, Fatima M, Zuliani J, Fontana R. Evaluation of the VITEK 2 system for identification and antimicrobial susceptibility testing of medically relevant gram-positive cocci. J Clin Microbiol, 2002; 40(5): 1681-6.

14. Spanu T, Sanguinetti M, Ciccaglione D, et al. Use of the VITEK 2 system for rapid identification of clinical isolates of staphylococci from bloodstream infections. J Clin Microbiol. 2003; 41(9): 4259-63.

15. Delport JA, Peters G, Diagre D, Lannigan R, John M. Identification of coagulase-negative staphylococci by the bruker MALDI-TOF biotyper compared to the Vitek 2 and MIS gas liquid chromatography. J Bacteriol Mycol Open Access. 2015; 1(1): 1-5.

16. Kloos WE, Zimmerman RJ, Smith RF. Preliminary studies on the characterization and distribution of Staphylococcus and Micrococcus species on animal skin. Appl Environ Microbiol. 1976; 31(1): 53-9.

17. Oeding P, Digranes A. Classification of coagulase-negative staphylococci in the diagnostic laboratory. Acta Pathol Microbiol Scand B. 1977; 85(2): 136-42.

18. Lamers RP, Muthukrishnan G, Castoe TA, Tafur S, Cole AM, Parkinson CL. Phylogenetic relationships among Staphylococcus species and refinement of cluster groups based on multilocus data. BMC Evol Biol. 2012; 12: 171.

19. D’Azevedo PA, Souza AG, Santos AF, Sales T, Neto TC, Pignatari AC. Suscetibilidade à novobiocina na identificação de amostras de Staphylococcus coagulase negativos (SCoN) isolados de hemoculturas. Rev Bras Anal Clin. 2007; 39(4): 303-4.

20. McTaggart LA, Elliott TSJ. Is resistance to novobiocin a reliable test for confirmation of the identification of Staphylococcus saprophyticus? J Med Microbiol. 1989; 30: 253-66.

21. Silveira ACO, D’Azevedo PA. Staphylococcus lugdunensis: um olhar diferenciado no laboratório clínico. J Bras Patol Med Lab. 2011; 47(2): 151-6.

22. Dupont C, Sivadon-Tardy V, Bille E, et al. Identification of clinical coagulase-negative staphylococci, isolated in microbiology laboratories, by matrix-assisted laser desorption/ionization-time of flight mass spectrometry and two automated systems. Clin Microbiol Infect. 2010; 16(7): 998-1004.

23. Carpaij N, Willems JL, Bonten MJM, Fluit AC. Comparison of the identification of coagulase-negative staphylococci by matrix-assisted laser desorption ionization time-of-flight mass spectrometry and tuf sequencing. Eur J Clin Microbiol Infect Dis. 2011; 30: 1169-72.

24. Loonen AJM, Jansz AR, Bergland JNB, Valkenburg M, Wolffs PFG, van den Brule AJC. Comparative study using phenotypic, genotypic, and proteomics methods for identification of coagulase-negative staphylococci. J Clin Microbiol. 2012; 50: 1437-9.