Comparison of Chromogenic Media to BD GeneOhm Methicillin-Resistant Staphylococcus aureus (MRSA) PCR for Detection of MRSA in Nasal Swabs

Abstract

To select a method for detecting methicillin-resistant Staphylococcus aureus (MRSA) in nasal swabs, we compared BD GeneOhm MRSA PCR and various culture media (mannitol salt agar with cefoxitin, MRSASelect, CHROMagar MRSA, and Spectra MRSA). While PCR detection of MRSA was more rapid, MRSASelect and Spectra MRSA demonstrated performance equivalent to that of PCR with maximal detection at 24 h.

The most significant risk factor for development of a methicillin-resistant Staphylococcus aureus (MRSA) infection is hospitalization (6), and colonization of the nares by S. aureus is a risk factor for development of MRSA and methicillin-susceptible Staphylococcus aureus infections (10, 14). Our institution currently performs active nasal surveillance for MRSA in surgical patients in the intensive care unit and is instituting a broader surveillance program. We assessed the performance of four different culture media and a molecular diagnostic test for the detection of MRSA in nasal swabs in an attempt to find an approach that would perform effectively from both analytical and work flow perspectives.

MRSA surveillance swabs of the anterior nares were collected on liquid Amies dual Bacti-swabs (Remel; Thermo Fisher Scientific, Lenexa, KS) by the care teams in the patient's unit. For the first part of the study, both swabs were inoculated onto mannitol salt agar containing cefoxitin (5 μg/ml) (MSA-FX) (12) and then one swab was removed and vortexed in 1 ml of saline. Fifty microliters of the saline suspension was aliquoted onto each of the following culture media: CHROMagar MRSA (BBL; Becton Dickinson, Franklin Lakes, NJ), MRSASelect (Bio-Rad, Hercules, CA), Spectra MRSA (Thermo Fisher Scientific), and MSA-FX. Plates were incubated in the dark at 35°C. Twenty-eight MRSA-positive cultures were detected by direct plating of swabs to MSA-FX, and 29 MRSA-positive cultures were detected by at least two plates following swab extraction in saline, so this process did not lead to reduced detection. Fifty microliters of the saline suspension was added to 200 μl of achromopeptidase (1 U/μl) in 10 mM Tris-HCl (pH 8.0), 1 mM EDTA buffer (Sigma-Aldrich, St. Louis, MO) for use in BD GeneOhm MRSA PCR (BD Diagnostics, San Diego, CA). For the second part of this study, one swab of the pair was inoculated onto MRSASelect medium, and the plates were read at 24 h. The other swab was added to a 1.5-ml tube. Either 200 μl or 300 μl of achromopeptidase (1 U/μl) was added, and the tube was vortexed for 10 s. Achromopeptidase samples from each part of the study were incubated at 37°C for 15 min, 99°C for 5 min, and on ice for at least 10 min. For BD GeneOhm MRSA PCRs, 2.8 μl of the lysate was used as template, and the assay was performed by following the manufacturer's protocol. Thermal cycling was performed on a Cepheid SmartCycler (Sunnyvale, CA). Detection of the mecA gene by PCR with agarose gel analysis of isolated colonies was performed as described previously (11).

Detection of MRSA by culture.

In part one of the study (370 specimens), there were 29 positive MRSA cultures (7.8%) (Table 1). MRSASelect and Spectra MRSA detected all of these positive specimens by 24 h. To determine the specificities of these media for MRSA detection, colonies on each plate with color consistent with MRSA were tested to determine if they were both coagulase positive (slide method) and contained the mecA gene by PCR. The false-positive (correct color on medium but coagulase negative and/or mecA negative) rates for the media were 0% for MRSASelect, 0.3% for Spectra MRSA, 2.1% for CHROMagar MRSA, and 6.5% for MSA-FX.

TABLE 1.

Comparison of media for detection of MRSA

Culture medium	No. of positive cultures (detection rate [%])a at time of incubation (h)
	18	24	48
MSA-FX	11 (38)	21 (72)	28 (97)
CHROMagar MRSA	8 (28)	26 (90)	27 (93)
MRSASelect	17 (59)	29 (100)	29 (100)
Spectra MRSA	17 (59)	29 (100)	29 (100)

^aTwenty-nine positive cultures were detected overall from 370 specimens. The detection rate was calculated as the number that were positive out of these 29.

Detection of MRSA with BD GeneOhm PCR using nasal swabs extracted in saline.

The same 370 specimens were analyzed in parallel using BD GeneOhm MRSA PCR (Table 2). The PCR result was positive for MRSA in 28 specimens (7.6%), including 24 specimens that were positive by culture. PCR was positive for MRSA in four specimens that did not contain MRSA by culture. For the five culture-positive specimens that did not have PCR-detectable MRSA, one of the residual swabs from each patient was then extracted directly in achromopeptidase. PCR was repeated on these lysates and was positive for all five specimens. In addition, the number of MRSA colonies for these specimens was less than 10 per plate for four of the specimens (data not shown). The four specimens that were positive for MRSA by PCR but negative by culture were all obtained from different patients. All four of these patients had received antibiotics prior to the nasal swab collection, but only two of them received antibiotics that could target MRSA (data not shown). None of these four patients had MRSA isolated from another site.

TABLE 2.

Comparison of BD GeneOhm MRSA PCR to culture for detection of MRSA in nasal swabs initially extracted in saline^a

PCR result	No. of specimens with culture result		Total no. of specimens
	Positive	Negative
Positive	24	4	28
Negative	5	337	342
Total	29	341	370

^aThe following determinations for PCR used MRSASelect and Spectra MRSA as the reference assays, as they demonstrated maximum detection: sensitivity, 24/29 (83%); specificity, 337/341 (99%); positive predictive value, 24/28 (86%); and negative predictive value, 337/342 (99%).

Inhibition of BD GeneOhm MRSA PCR is affected by the swab extraction method.

The effects of the following three methods of nasal swab extract preparation on PCR inhibition were compared: (i) vortexing a swab in 1 ml of saline and then removing 50 μl of this suspension to 200 μl of achromopeptidase solution; (ii) adding 200 μl of achromopeptidase directly to one nasal swab; and (iii) adding 300 μl of achromopeptidase to a swab (Table 3). A freeze-thaw of the extract resolved all the inhibited samples in the saline extraction group and all but three in the group prepared directly in 300 μl of achromopeptidase (Table 3). Preparation of the swab directly in 200 μl of achromopeptidase yielded a significantly higher rate of inhibition in our hands (43%) than previously described (11.7%) (8). This difference could be accounted for by technical experience or the presence of a PCR inhibitor potentially arising from differences in collection procedures. This latter possibility is consistent with dilution of the inhibitor by the increase of the extraction volume to 300 μl.

TABLE 3.

Rates of inhibition of BD GeneOhm MRSA PCR following different methods of swab extraction

Sample preparationa	No. of samples	No. (%) of samples showing:
		PCR inhibition	PCR inhibition after freeze-thawb
Saline	370	18 (4.9)	0 (0)
200 μl ACP	142	61 (43.0)	ND
300 μl ACP	327	52 (15.9)	3 (0.9)

^aThe extracts for PCR were prepared by vortexing the swab in saline and adding an aliquot to 200 μl ACP, adding 200 μl ACP directly to the swab, or adding 300 μl ACP directly to the swab. ACP, achromopeptidase.

^bSamples initially inhibited were frozen at −80°C for at least 24 h, and PCR was then repeated.

Comparison of MRSASelect to BD GeneOhm MRSA PCR for detection of MRSA in nasal swabs.

In part two of the study, we compared MRSASelect plates to BD GeneOhm MRSA PCR using swabs extracted in 300 μl of achromopeptidase (327 specimens) (Table 4). MRSASelect cultures detected MRSA in 23 samples (7.0%), while BD GeneOhm MRSA PCR detected MRSA in 24 samples (7.3%). Two specimens were positive for MRSA by MRSASelect but negative by BD GeneOhm MRSA PCR. Both of these cultures had a single colony that was consistent with MRSA, both of which were confirmed to be MRSA by the coagulase test and mecA PCR.

TABLE 4.

Comparison of BD GeneOhm MRSA PCR using swabs extracted in 300 μl of achromopeptidase to MRSASelect culture^a

PCR result	No. of specimens with culture result		Total no. of specimens
	Positive	Negative
Positive	21	3	24
Negative	2	298	300
Total	23	301	324

^aThe following determinations for PCR used MRSASelect as the reference assay: sensitivity, 21/23(91%); specificity, 298/301 (99%); positive predictive value, 21/24 (88%); and negative predictive value, 298/300 (99%). The three samples that remained PCR inhibited after the freeze-thaw were excluded from this analysis.

Three specimens were positive for MRSA by BD GeneOhm PCR but negative on MRSASelect medium. S. aureus was also not detected upon restreaking these swabs to sheep blood agar. The first specimen was obtained from a swab taken on a different day from one of the patients who had a single MRSA colony on MRSASelect but a negative PCR result. This patient had MRSA isolated from an elbow culture and had received vancomycin. The patient had a total of four MRSA screening nasal swabs, two that had the discrepant PCR and culture results and two that were negative by both MRSASelect culture and BD GeneOhm PCR. For the second of the PCR-positive, culture-negative specimens, the patient had MRSA cultured from blood and peritoneal fluid and was started on vancomycin and ceftriaxone one day prior to obtaining the nasal swab. The third patient for whom the specimen was PCR positive and culture negative had a repeat swab the following day that was negative by both PCR and culture. This patient did not have MRSA isolated from any additional source and did not receive antibiotics.

Summary.

Our results indicate that both MRSASelect and Spectra MRSA provide culture-based methods to detect MRSA in 24 h with high detection and low false-positive rates without the need for additional subculturing. Prior studies of MRSASelect have reported sensitivities of 63% and 97% at 18 and 24 h, respectively (9), 99% at 24 h (1), 97% at 18 h (12), and 80% at 42 h (7). The latter study used an enrichment broth step for comparison, which may partially explain the lower sensitivity. Two of these studies also included CHROMagar MRSA medium, and in both cases, MRSASelect was found to be more sensitive (7, 12). In a separate analysis, the sensitivity of CHROMagar MRSA was higher at 48 h (95%) than at 24 h (87%), which is consistent with our results (5). To our knowledge, this study is the first published report describing the performance of Spectra MRSA.

Compared to the performance of chromogenic media (MRSASelect and Spectra MRSA in the first part of the study and MRSASelect in the second part of the study), BD GeneOhm MRSA PCR was 83% and 91% sensitive for saline eluates and direct swabs, respectively, and was 99% specific in both parts of the study. Prior studies analyzing this assay reported sensitivities ranging from 86 to 100% and specificities ranging from 89 to 99% (2-4, 13, 15). In our study, PCR may have detected MRSA carriage not detected by culture in patients receiving antibiotics targeting MRSA, but the clinical significance of this finding is unknown. While BD GeneOhm PCR can be completed in 2 h, the requirement for specimen batching would likely lengthen this turnaround time. Our study does not demonstrate a clear analytical advantage for MRSASelect, Spectra MRSA, or BD GeneOhm PCR, so additional laboratory-specific factors, including financial considerations, turnaround time, and technical expertise, will be important in deciding between these screening methods.