Indian Journal of Medical Microbiology Medknow Publications on behalf of Indian Association of Medical Microbiology ISSN: 0255-0857 EISSN: 1998-3646 Vol. 29, Num. 1, 2011, pp. 42-46

Indian Journal of Medical Microbiology, Vol. 29, No. 1, January-March, 2011, pp. 42-46

Original Article

Comparison of four methods for rapid identification of Staphylococcus aureus directly from BACTEC 9240 blood culture system

NS Ozen¹, D Ogunc¹, D Mutlu¹, G Ongut¹, BO Baysan¹, F Gunseren²

¹ Department of Clinical Microbiology, Akdeniz University Medical Faculty, Antalya, Turkey
² Department of Clinical Microbiology and Infectious Diseases, Akdeniz University Medical Faculty, Antalya, Turkey

Correspondence Address: D Ogunc, Department of Clinical Microbiology, Akdeniz University Medical Faculty, Antalya, Turkey, dogunc@akdeniz.edu.tr

Date of Submission: 05-Feb-2010
Date of Acceptance: 08-Jul-2010

Code Number: mb11010

PMID: 21304194

DOI: 10.4103/0255-0857.76523

Abstract

Keywords: Blood culture system, direct identification, polymerase chain reaction, S. aureus, slide agglutination test

Introduction

The correct identification and differentiation of Staphylococcus aureus (S. aureus) from Coagulase-Negative Staphylococci (CoNS) is important for patient mortality and morbidity. ^[1] S. aureus is always considered to represent a true bacteremia when isolated from a blood culture. On the other hand, CoNS often represent contamination in a significant proportion of blood cultures.^[2] However, it is also important to confirm the presence of CoNS in clinical samples since the incidence of infections caused by this organism is on the rise. ^[3] The full identification of S. aureus usually takes 24 h by using the traditional methods.^[4],[5] S. aureus is differentiated from CoNS most reliable by tube coagulase test (TCT), which is an inexpensive method in routine laboratory practice.^[6] Multiple studies have evaluated the use of rapid tests for detection of S. aureus directly from blood cultures. TCT has been reported to be highly specific with varying sensitivities.^{[7],[8],[9],[10]} Other alternative methods are slide agglutination tests for clumping factor, protein-A, and other surface antigens. ^[6] However in recent reports, these tests showed poor sensitivities when performed directly from blood culture broth (BCB) and it is recommended that negative results should be confirmed. ^[8],[11] Several nucleic acid based amplification methods have been reported for the identification of S. aureus. The same-day differentiation is the advantage of molecular tests but they are expensive and needs qualified personnel.^[2],[5],[12] Polymerase chain reaction (PCR) from BCB has been reported to be highly predictive of S. aureus in most studies. ^[13],[14] Some commercial real-time PCR kits have been developed to differentiate S. aureus and/or CoNS from BCB. In a study, the LightCycler Staphylococcus M ^Grade kit was evaluated to differentiate S. aureus and CoNS from BacT/Alert bottles and showed sensitivity and specificity of 100 and 98.44%, respectively.^[15] To our knowledge, there is no study with Dry Spot Staphytect Plus and LightCycler Staphylococcus M ^Grade kit, which was performed directly from BACTEC-9240 system. In this study, we evaluated the performance of TCT, slide agglutination test (Dry Spot Staphytect Plus), conventional PCR and LightCycler Staphylococcus M ^Grade kit directly from blood culture bottles to achieve rapid identification of S. aureus by using the BACTEC-9240 blood culture system.

Materials and Methods

A total of 129 BCBs from February to June 2006 were included in the study. Bottles that gave a positive signal in the BACTEC-9240 blood culture system were removed, and a gram stain was performed. Broths that revealed gram-positive cocci suggesting Staphylococci were included in the study. Standard laboratory procedures including inoculation onto 5% sheep blood agar, incubation at 37ºC, catalase and coagulase test were used to identify gram-positive cocci. A multiplex PCR targeting a staphylococcus-specific region of the 16S rRNA gene, PBP2A gene (mecA) and S. aureus-specific thermostable nuclease gene (nuc) was used for identification of staphylococci from bacterial colonies. This method was accepted as the reference method.

Direct tube coagulase test

The test was performed as described by McDonald et al.^[9] briefly; 10 ml of BCB was removed from the bottles and three drops were used for TCT test.

Pelleted tube coagulase test

The supernatant of the remaining BCB was transferred to the second tube after centrifugation at 150× g for 10 minutes. The pellet was obtained from the supernatant after centrifugation at 1000× g for 15 minutes and was resuspended in 1 ml of 0.85% sterile saline and a three-drop sample was tested.

Clot formation was examined for direct and pelleted TCT at two and four hours at 35ºC and 24 h after incubation at room temperature.

Direct slide agglutination test

The test (Dry Spot Staphytect Plus) was performed according to the manufacturer′s instructions, except that one drop of BCB was used instead for bacterial growth suspension.

Pelleted slide agglutination test

The test was performed according to the manufacturer′s instructions. One drop of the pellet suspension as described above in the pelleted TCT was used.

Conventional Polymerase chain reaction from blood culture broth

DNA was extracted by using the protocol described previously by Louie et al.^[13] Briefly, 100 μl of BCB was transferred to 1 ml of distilled water and left for five minutes at room temperature. The mixture was centrifuged at 16000× g for one minute and supernatant was removed. The pellet was resuspended in 100 μl Triton X-100 lysis buffer. Five microliter of lysostaphin (100 mg/ml) was added to this suspension and incubated for 10 minutes at 37ºC. Then the suspension was boiled for 10 minutes and left for 5 minutes at room temperature. The suspension was centrifuged at 16000× g for one minute. One microliter of the supernatant was used as template DNA for multiplex PCR. A multiplex-PCR targeting 16S rRNA gene, mecA gene and the nuc gene was performed for the detection of methicillin resistance and identification of S. aureus.

Conventional Polymerase chain reaction from bacterial colonies

The DNA extraction procedure was based on a modification by Merlino et al.^[16] Approximately 4 McFarland bacterial suspension was prepared from culture plates and centrifuged for 5 minutes at 14000× g. After the supernatant was removed, 100 μg/ml of lysostaphin was added and incubated at 37ºC for 10 minutes. About 100 μg/ml of proteinase-K was added to the mixture and incubated at 37ºC for 20 minutes. Later, proteinase-K was inactivated by heating for ten minutes at 90ºC and was left at room temperature for five minutes, and then centrifuged for five minutes at 14000× g. Supernatant was used as a template for multiplex PCR. All the amplified PCR products were screened on 2% agarose gel.

The same primers were used in conventional PCR from BCB and conventional PCR from bacterial colonies, as shown in [Table - 1].

LightCycler assay

Real-time PCR was performed in the LightCycler instrument according to the manufacturer′s instructions. The LightCycler reaction protocol was as follows: denaturation and Taq polymerase activation [95ºC for 10 minutes], 45 cycles of PCR (1 cycle consists of 95ºC for 10 seconds, 50ºC for 15 seconds, and 72ºC for 10 seconds), a melting phase (95ºC for 60 seconds and 40 to 80ºC in 60 seconds) and a cooling phase at 40ºC for 30 seconds. Melting curve analysis was performed for all samples with a positive amplification curve with fluorescence channel F2. The presence of PCR inhibitors is detected with fluorescence channel F3 for all samples. Samples that have Tm values of 62.1 ± 2ºC and Tm values between 45.4 ± 2 and 57.5 ± 2ºC were interpreted as S. aureus and CoNS, respectively.

Staphylococcus aureus ATCC 29213 and Staphylococcus epidermidis ATCC 12228 were used for positive control in all tests.

Statistical analysis was done with McNemar test.

Results

A total of 129 bottles were studied of which 55 grew S. aureus and 69 grew CoNS. Five vials grew both S. aureus and CoNS.

Number of S. aureus and CoNS isolates identified by different tests are shown in [Table - 2].

Sensitivities, specificities, positive predictive values (PPV) and negative predictive values (NPV) of all performed tests are shown in [Table - 3].

Direct TCT correctly identified 57 (95%) of 60 S. aureus isolates. Three isolates were negative after incubation of 2, 4, and 24 h. Only one S. aureus isolate was misidentified with pelleted TCT. The sensitivity of pelleted TCT was higher than direct TCT but the difference was not statistically significant. There was no statistically significant difference between the TCT test results read at 2, 4, and 24 h.

Direct and pelleted slide agglutination test correctly identified 26 (44%) and 41 (68%) of 60 S. aureus isolates, respectively. The sensitivity of pelleted slide agglutination test was significantly higher than that of direct agglutination test (P < 0.005).

Results of conventional PCR assay for 16S rRNA gene, nuc and mecA detection performed from blood culture bottles were in agreement with those of reference method for 99% of the strains. Only one bottle which contained both S. aureus and CoNS was detected as CoNS.

All bottles containing S. aureus and/or CoNS were correctly identified by the LightCycler Staphylococcus M ^Grade kit.

Discussion

The need for rapid identification of S. aureus from blood cultures prompted the development of alternative diagnostic methods. Rapid identification and susceptibility testing could help the clinician for the adjustment of antibiotic treatment for the effective management of patients with staphylococcal bacteremia. ^[4]

In a study, sensitivities of 2 h TCT were found to be 86.2 and 84.4% for a direct and pelleted broth with a specificity of 100%, respectively. ^[9] In another study, it was reported that increasing the incubation period from 2 to 4 h enhanced sensitivity and suggested that the negative results must not be reported until 24 h. ^[17] Chopin and Musgnug. evaluated TCT for the rapid identification of S. aureus from BCB. They reported a sensitivity of 84.1% for direct TCT at 4 h and showed that sensitivity increased to 97% if the test was read at 24 h. ^[7] In another study, it was reported that 2 h TCT was highly predictive for the rapid identification of S. aureus from the blood culture pellets. They also found no significant difference between the test results read at 2, 6, and 24 h. Test sensitivities ranged from 87 to 92% for 2 and 24 h with specificities of 100%. ^[8] In our study the sensitivity of TCT was similar or slightly higher than those reported previously and the performance of the test results read at 2, 4, and 24 h were not statistically different.

Slide agglutination kits that have not been designed for direct use on BCB may have contributed to the low sensitivities of the tests used for the rapid identification of S. aureus. Previous studies have reported sensitivities of slide agglutination tests from BCB varying from 12 to 78%. ^[8] In our study, sensitivity of dry spot test from BCB was 44%. However, sensitivity increased to 68% when performed from pellet. The difference was statistically significant between direct and pelleted slide agglutination test (P < 0.002). On the other hand, the sensitivity of dry spot test from bacterial colonies was reported as 100% in a recent report. ^[18] Our data suggested that neither direct nor pelleted slide agglutination test was sufficiently predictive for use as a routine diagnostic test from BCB.

Non-culture-based determination of S. aureus and CoNS with molecular methods can be achieved 18-24 h earlier than conventional methods. ^[5] In a study by Tan et al., the sensitivity of conventional PCR directly performed from BACTEC bottles was found to be 96%. ^[14] We found the sensitivity of PCR 99%, which is similar with the study of Louie et al. ^[13] Only one bottle that contains both S. aureus and CoNS was misidentified as CoNS with PCR. PCR enables to detect S. aureus one day earlier than conventional methods. The sensitivity of PCR can be affected by presence of PCR inhibitors, personnel knowledge, and other physical factors. ^[5] Another limitation of PCR is that misidentifications can occur if both S. aureus and CoNS co-exist in the same bottle. ^[4],[19]

The LightCycler Staphylococcus M ^Grade kit can differentiate S. aureus from CoNS if both are present in a mixed culture. In a study, all isolates were correctly identified with LightCycler kit with a sensitivity and specificity of 100%. ^[15] In our study, LightCycler Staphylococcus M ^Grade kit correctly identified S. aureus and CoNS in a mixed culture, which was identified as CoNS with conventional PCR. The advantage of this technique over conventional PCR method is the ability to detect more than one Staphylococcus species in the same sample although there was limited number of mixed cultures in our study. Further studies with a larger sample size are needed to evaluate the promising role of LightCycler Staphylococcus M ^Grade kit for the accurate detection of S. aureus and CoNS in mixed blood cultures.

The use of PCR technology can be a good tool for rapid identification of S. aureus from BCB. LightCycler has excellent specificity and sensitivity but its cost limits its use in routine laboratories. Conventional PCR, which is a labor-intensive and relatively expensive method does not detect mixed cultures. ^{[4],[19],[20]}

Our findings suggest that TCT is a practical, rapid, cost-effective and sensitive method for the detection of S. aureus from BCB.

We suggest that use of the 2 h TCT directly from BCB is an excellent way to distinguish S. aureus from CoNS. It is inexpensive and simple to perform with high specificity and sensitivity when performed by trained personnel. ^[7],[8],[17]

We believe that traditional methods are still useful in many resource-limited laboratories.

Acknowledgments

This study was supported by Akdeniz University Scientific Research Unit.

References

Copyright 2011 - Indian Journal of Medical Microbiology