INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Despite guidelines published by the National Committee for Clinical Laboratory Standards (NCCLS) for the testing of susceptibility to oxacillin for staphylococci, the optimal phenotypic method for detecting methicillin (oxacillin) resistance remains controversial. The objective of the present study was to determine which of the following susceptibility test methods, performed by using recommended or modified NCCLS guidelines, best detected oxacillin resistance: agar dilution, disk diffusion, and agar screen (swab or spot inoculation). The results for these methods were compared to PCR detection of the mecA gene for 58 clinical isolates of coagulase-negative Staphylococcus spp. (CoNS) and 41 clinical isolates of Staphylococcus aureus.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Ninety-nine clinical isolates (58 CoNS isolates and 41 S. aureus isolates) and four control strains (S. aureus ATCC 25923 [lacking mecA], S. aureus MC 205 [a Mayo Clinic isolate lacking mecA], Staphylococcus epidermidis ATCC 27626 [mecA positive], and S. aureus MC 206 [a Mayo Clinic isolate, mecA positive]) were evaluated. All of the clinical isolates were obtained from human specimens submitted to the Mayo Clinical Microbiology Laboratory. No two isolates were from the same patient, and no isolates were part of nosocomial outbreaks. All staphylococcal isolates and ATCC control strains were screened for the presence of the mecA gene by using a modification of a previously described multiplex PCR method (6). The following PCR primers were used for amplification of the mecA gene: mec449F, 5'-AAA CTA CGG TAA CAT TGA TCG CAA C-3', and mec761R, 5'-CTT GTA CCC AAT TTT GAT CCA TTT G-3'. Primers specific to staphylococcus 16S rRNA, i.e., 16S 387F, 5'-CGA AAG CCT GAC GGA GCA AC-3', and 16S 914R, 5'-AAC CTT GCG GTC GTA CTC CC-3', were used in the multiplex PCR to provide a positive control for target amplification. The PCR mix contained 200 µM deoxynucleotide triphosphates, 10 mM Tris (pH 8.3), 50 mM KCl, 1.5 µM MgCl₂, 10% glycerol, 200 µM mec primers, 50 µM 16S rRNA primers, and 0.025 U of AmpliTaq DNA polymerase (PE Applied Biosystems, Foster City, Calif.) per µl. Target DNA (2 µl) was added to 48 µl of mix and then thermocycled for 30 cycles of 95°C for 30 s, 62°C for 30 s, and 72°C for 30 s. PCR amplicons were analyzed by gel electrophoresis. With this modification, a 313-bp fragment of the mecA gene and a 528-bp fragment of the 16S rRNA gene unique to staphylococci are amplified. By this analysis, 30 (52%) of 58 CoNS clinical isolates and 17 (42%) of 41 S. aureus clinical isolates were shown to carry the mecA gene.

For susceptibility testing, the information about the media used (including whether each medium was prepared in-house), oxacillin concentration, incubation parameters, and susceptibility interpretive guidelines is shown in Table 1. Recent studies suggest that oxacillin susceptibility testing of CoNS isolates at lower breakpoints may more reliably detect oxacillin resistance encoded by the mecA gene (5, 17, 19, 24). Therefore, as part of the agar screen evaluation for CoNS isolates, we used a Mueller-Hinton (MH) plate containing 0.6 µg of oxacillin per ml in addition to a standard MH plate containing 6.0 µg of oxacillin per ml. Also, the results for the agar dilution were interpreted with 1-µg/ml concentration of oxacillin regarded as the breakpoint for susceptibility (analysis 2, Table 1).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Results for CoNS isolates are displayed in Table 2, and those for S. aureus isolates are displayed in Table 3. Based on the ability to support growth and the results for susceptibility testing, the best phenotypic methods for detecting mecA gene-encoded oxacillin resistance for CoNS isolates were agar dilution by using MH agar (Difco) supplemented with 4% NaCl and incubation at 35°C in ambient air for 48 h (there were no growth failures, and sensitivity was 96.7%) and agar screen (swab inoculation) by using MH medium (Difco) prepared in-house supplemented with 4% NaCl and containing 0.6 µg of oxacillin/ml (there was one growth failure, and sensitivity was 100%). The single CoNS isolate that failed to grow for the agar screen plate did not carry the mecA gene. This isolate grew poorly or not at all for all methods tested. The 96.7% sensitivity for the former method resulted from one very major error; one isolate that carried the mecA gene was interpreted as susceptible regardless of whether a susceptibility breakpoint of 2 µg/ml or of 1 µg/ml was used. However, when this isolate was evaluated by the agar screen by using the swab inoculation, 20 colonies grew on the plate with oxacillin concentration of 6.0 µg/ml and 60 colonies grew on the plate with oxacillin concentration of 0.6 µg/ml.

All methods, with the exception of one (agar dilution without added NaCl with incubation at 30°C for 48 h), correctly identified all S. aureus isolates with mecA-encoded oxacillin resistance; no growth failures occurred with any method. Varying the temperature for incubation (30 or 35°C) had little effect on results for both S. aureus and CoNS isolates.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Numerous studies have been conducted to determine optimal methods for phenotypic detection of oxacillin (methicillin) resistance among clinical isolates of staphylococci. Table 4 summarizes a number of these studies; only recent reports for commercial (including automatic) systems are included as many of the early reports indicated inferior performance for these methods. Oxacillin (methicillin) resistance for S. aureus isolates can be reliably detected by a variety of phenotypic methods. As shown in Table 4, several studies have demonstrated that 100% of oxacillin (methicillin)-resistant S. aureus test isolates were detected by either broth dilution (20), agar dilution (6, 8, 22), agar spot screen (9, 20), gradient diffusion (Epsilometer test) (22), or disk diffusion (4, 8, 12, 21) method or by the automated API-Plus system (bioMerieux) (21). For these methods, different concentrations of NaCl in media (0 to 4%) or varying incubation times (24 or 48 h) had little effect.

In contrast, NaCl supplementation, incubation time, and in one instance, inoculum were important parameters for expression of oxacillin (methicillin) resistance for CoNS isolates. As shown in Table 4, detection of oxacillin (methicillin) resistance was best achieved (sensitivity, 97 to 100%) if the NaCl concentration in medium was 4 to 5% (4, 20, 22, 24), the incubation time was 48 h rather than 24 h (4, 17, 22, 24), and/or a larger inoculum was used (14). Furthermore, when these parameters were applied, non-broth-based methods (agar spot [4, 20] or swab [4, 24] screens, disk diffusion [14, 17], or gradient diffusion [Epsilometer test] [4, 22]) had the highest accuracy. Of interest, two studies reported by the same group of investigators concluded that 2%, and not 4%, NaCl supplementation produced the best results for phenotypic detection of oxacillin resistance when agar dilution, broth dilution, and gradient diffusion testing were compared (1, 10). The same group of organisms was used in each study; however, the reference standards were broth microdilution (MH medium with 2% NaCl and incubation for 24 h at 35°C) for one study (1) and identification of the mecA gene for the other study (10). However, for neither of these studies was the incubation period extended beyond 24 h.

The results of our evaluation for S. aureus isolates are in agreement with those of many of the studies summarized in Table 4. That is, mecA gene-associated resistance is reliably detected by a variety of phenotypic methods for which varying NaCl supplementation or incubation time has little effect. In our study, a 24-h incubation period was sufficient for mecA-associated resistance in S. aureus isolates for all of the methods we tested; extending the incubation time to 48 h frequently resulted in decreased specificities. These decreases in specificity were minor and may have occurred as the result of a decrease in the bioactivity of antimicrobial in the test media over time but could relate to mechanisms associated with oxacillin (methicillin) resistance not involving the mecA gene. These mechanisms might include hyperproduction of -lactamase, production of penicillin binding proteins other than PBP2a encoded by mecA, which have decreased affinity for methicillin or related compounds, enzymes which inactivate methicillin, or as yet undiscovered mechanisms (3). Geha and colleagues (6) noted that among 228 clinical S. aureus isolates, 44 were methicillin resistant by agar dilution and disk diffusion techniques. Forty of these isolates carried the mecA gene as assessed by PCR; three of the remaining four isolates were demonstrated to be hyperproducers of -lactamase. Kolbert and colleagues (15) noted that among 147 consecutive clinical S. aureus isolates, 28 were resistant by using a disk diffusion method. Fourteen of those isolates possessed the mecA gene; the remaining 14 isolates did not carry the mecA gene and were felt to be hyperproducers of -lactamase.

The results of our study for CoNS isolates corroborate the results of several studies summarized in Table 4 in which similar media and incubation times were used. That is, oxacillin (methicillin) resistance encoded by the mecA gene, by an agar-based method, is best detected by using MH medium supplemented with 4% NaCl and incubation for 48 h (4, 20, 22, 24).

Current NCCLS recommendations for oxacillin susceptibility testing of staphylococci by using the agar dilution method specify the use of MH medium supplemented with 2% NaCl and incubation at 35°C in ambient air for 24 h (18). In contrast, NCCLS recommendations indicate that the oxacillin agar screen method should be used only for S. aureus isolates and that for this method, MH medium should be supplemented with 4% NaCl (not 2% NaCl) and incubation should be in ambient air for 24 h (18). The findings of the present study indicate that the medium specified for the agar screening method (MH medium with 4% NaCl) should also be used for the agar dilution method. Furthermore, the present study supports the use of the agar screen plate for CoNS isolates as well as S. aureus isolates.

In our study, extension of the incubation period to 48 h for CoNS, but not for S. aureus isolates, improved sensitivity regardless of the test method used. Like S. aureus isolates, specificity decreased for CoNS isolates when the incubation period was extended to 48 h and was most pronounced for disk diffusion methods. Like S. aureus isolates, CoNS isolates that are mecA negative but phenotypically resistant to oxacillin may possess other mechanisms for resistance. Geha and colleagues (6) noted that among 272 CoNS isolates, 148 were methicillin resistant and all of these possessed the mecA gene. However, Kolbert and associates (15) noted that among 253 consecutive clinical CoNS isolates, 128 of 130 mecA-positive isolates were resistant by disk diffusion. Thirteen additional isolates that were resistant by disk diffusion did not possess mecA, and three of these were shown to be -lactamase hyperproducers.

We also observed that for agar screen methods and testing of CoNS isolates, media prepared in-house performed better than commercially prepared media. All commercially prepared media were used for testing prior to the expiration dates. We are unsure as to the reasons for the better performance of media prepared in-house. Unlike for media prepared in-house, we were unable to determine growth failures due to the unavailability of growth control plates (i.e., plates, provided by the manufacturer, that do not contain oxacillin). Undetected growth failures would have been misinterpreted as indicating susceptibility.

It has been suggested that the growth of heteroresistant subpopulations of staphylococci may be enhanced by using a cooler incubation temperature (i.e., 30 rather than 35°C) (13). However, our results showed that varying the temperature of incubation from 30 to 35°C had little effect.

The oxacillin susceptibility breakpoints currently recommended by the NCCLS for dilution testing methods are 2 µg/ml for S. aureus and <0.25 µg/ml for CoNS. The lower breakpoint for CoNS (compared with that for S. aureus) is a recent recommendation (18). Our study supports this recommendation. Specifically, we have demonstrated that for agar dilution and testing of CoNS isolates (analysis 2, Table 1), a lower susceptibility breakpoint of 1 µg of oxacillin/ml (instead of 2 µg of oxacillin/ml) permitted the detection of more CoNS isolates with mecA-associated resistance (Table 2).

To our knowledge, our study is the first to evaluate lower breakpoints for an agar screen method. Indeed, if the breakpoint for susceptibility for CoNS isolates was decreased 10-fold, from 6 µg/ml to 0.6 µg/ml, 100% sensitivity was achieved. York and colleagues concluded that lowering the breakpoint for susceptibility to less than 2 µg/ml would increase the sensitivity of the broth microdilution method when CoNS isolates are tested (24). Mulder demonstrated that oxacillin resistance was best detected by the E-test when less than 2 µg of oxacillin/ml was used as the breakpoint (17). Frebourg and colleagues also noted that a decrease in the breakpoint for oxacillin susceptibility should improve results for staphylococcal testing for both the E-test and the Vitek method (5). Finally, Ramotar and colleagues recently reported that among 188 CoNS clinical isolates reported as susceptible by automated methods, 16 were positive for the mecA gene by PCR analysis (19). For two of these isolates MICs of oxacillin were equal to 1 µg/ml, and for four isolates the MICs were 0.5 µg/ml (19). The current study suggests that a breakpoint near 0.5 µg of oxacillin/ml is more reasonable for testing CoNS isolates whether agar dilution or agar screening plates are used. As previously mentioned, current NCCLS guidelines advise use of the agar screening plate for S. aureus isolates but not CoNS isolates. We believe that the agar plate can be useful for detecting oxacillin resistance for CoNS isolates but that a lower concentration of oxacillin should be used for testing CoNS isolates than that used for S. aureus isolates.

In conclusion, we recommend that the following parameters for the agar dilution or agar screening methods be used for testing of CoNS isolates: agar dilution by using MH medium supplemented with 4% NaCl with incubation at 35°C for 48 h, and agar swab screen by using MH medium supplemented with 4% NaCl (prepared in-house) with incubation at 35°C for 48 h. The breakpoint for susceptibility should be 0.5 µg/ml.