Performance of Fusidic Acid (CEM-102) Susceptibility Testing Reagents: Broth Microdilution, Disk Diffusion, and Etest Methods as Applied to Staphylococcus aureus

Staphylococcus aureus is a leading cause of skin and skin structure infections (SSSI), hospital- and community-acquired bacterial pneumonia, and nosocomial bloodstream infections (BSI) (3, 9, 14). Resistance to methicillin (oxacillin) among S. aureus (MRSA) isolates ranges from 30% to more than 60% and is present worldwide (9, 10, 17). Furthermore, hospital-associated strains of MRSA (HA-MRSA) are often multidrug-resistant (MDR), exhibiting resistance to all β-lactam agents, penems, carbapenems, aminoglycosides, macrolides, tetracyclines, trimethoprim, and fluoroquinolones (3). The emergence of MRSA strains with reduced vancomycin susceptibility further reduces treatment options (4). Although community-associated MRSA (CA-MRSA) presently remains susceptible to clindamycin, most tetracyclines, and trimethoprim-sulfamethoxazole, its emergence as a cause of infection in health care facilities is a growing source of concern (19). These resistance issues associated with such a virulent and prevalent pathogen have spurred the development of new antistaphylococcal agents, as well as reconsideration of the role of older agents with demonstrated antistaphylococcal activity (4, 10). Regarding MDR-MRSA, it has been suggested that the use of an agent such as fusidic acid may prove useful in treating these difficult infections and could help to delay the inevitable development of resistance to newer agents, such as linezolid and daptomycin (15). A promising feature of fusidic acid is the lack of cross-resistance with other antimicrobial classes as a result of the unique mode of action that inhibits bacterial protein synthesis at the translational stage (2, 12, 15).

Although fusidic acid has been used throughout much of the world for more than 40 years (1, 15), U.S. Food and Drug Administration (FDA) licensure has never been obtained, and the drug is not currently available in the United States. As a result, resistance to fusidic acid is extremely uncommon among U.S. strains of S. aureus, including methicillin-susceptible, MRSA, and vancomycin-intermediate and -resistant strains, as well as those strains with decreased susceptibility to linezolid and daptomycin. Despite the fact that in vitro susceptibility testing of fusidic acid has been performed for many years, fusidic acid is not presently included in the tables of the Clinical and Laboratory Standards Institute (CLSI), and interpretive breakpoints for MIC and disk diffusion testing of fusidic acid against S. aureus are not available (11, 20).

Previous authors have demonstrated that susceptibility of staphylococci to fusidic acid may be indicated at MICs of ≤0.25, ≤0.5, or ≤1 μg/ml and resistance at MICs of ≥2 μg/ml (11, 15, 20). Recently, Skov et al. (20) utilized CLSI reference broth microdilution and disk diffusion methods to propose staphylococcal susceptibility interpretive criteria of ≤0.5 μg/ml (≥21 mm) and resistance criteria of ≥2 μg/ml (≤18 mm). The EUCAST (13) organization has selected ≤1 μg/ml as susceptibility breakpoint for MIC testing. In the present study, we provide additional fusidic acid MIC and disk diffusion data to support the findings of Skov et al. (20) or EUCAST (13) and, in addition, evaluate the utility of the Etest (AB Biodisk, Solna, Sweden) methodology for testing this agent against a large North American collection of S. aureus strains (13, 20).

A total of 778 nonduplicate clinical isolates of S. aureus (52% MRSA) from patients with SSSI or BSI were obtained from more than 30 medical centers in the United States and Canada between 1997 and 2006. A subset of CA-MRSA isolates (50 strains from the United States) were tested as a resistance subset only. All isolates (778 overall) were forwarded to the monitoring laboratory (JMI Laboratories, North Liberty, IA) for subsequent identification confirmation and reference antimicrobial susceptibility testing. Identification was performed using an automated system (Vitek; bioMerieux, Hazelwood, MO) or conventional manual methods, as required.

All strains were tested by the CLSI broth microdilution method using prepared and validated frozen-form panels in cation-adjusted Mueller-Hinton broth (6). Fusidic acid (also known as CEM-102; Cempra) reference powder was obtained from Cempra Pharmaceuticals (Chapel Hill, North Carolina). Disk diffusion testing (all strains) was performed according to the CLSI method (5) using Mueller-Hinton agar and two disk concentrations (5 [728 strains] and 10 μg [778 strains]). The zone diameters were measured to the nearest mm using a caliper (Fig. 1a). Etest was performed as recommended by the manufacturer (AB Biodisk) using Mueller-Hinton agar, with inoculums of 1 × 10⁸ to 2 × 10⁸ CFU/ml (5) and incubation at 37°C in air for 18 to 24 h. The MIC was read at 80% inhibition relative to the growth of the control. The organism collection (excluding the CA-MRSA subset) (728 strains) was used to directly compare the two disk tests and the reference versus Etest MIC results (Fig. 1, 2, and 3). All fusidic acid-nonsusceptible strains (14 total) were found with each test method.

Quality control (QC) was performed concurrently with all testing determinations, using S. aureus ATCC 29213 (MIC) or ATCC 25923 (disks) and S. pneumoniae ATCC 49619. The proposed QC ranges for MIC and disk diffusion (10 μg) tests for S. aureus ATCC 29213 and ATCC 25923 were 0.06 to 0.25 μg/ml and 24 to 32 mm, respectively. The ranges for S. pneumoniae ATCC 49619 were 4 to 32 μg/ml and 8 to 16 mm, respectively (16). Among 61 replicates, all QC values were within control ranges (8, 16).

Broth microdilution test results were compared to zone diameters of inhibition around 5- and 10-μg fusidic acid disks by scattergram analysis and regression line equations. Interpretive zone size criteria were established using the error rate-bounded method of Metzler and DeHaan (18) as described by CLSI document M23-A3 (7). Correlation between the MIC methods (broth microdilution and Etest) was performed by scattergram and regression analysis. The essential agreement between the two methods was calculated, as well as the percentage of results within plus-or-minus one log₂ dilution step, optimized to 95% (7).

Among strains of S. aureus tested in this study, 14 were resistant to fusidic acid as defined by a breakpoint of ≥2 μg/ml (Fig. 1a). Excellent correlation (r = 0.74) was noted between broth microdilution MICs and zone diameters obtained with the 10-μg disk test (Fig. 1a). Using a susceptible MIC breakpoint of ≤0.5 or ≤1 μg/ml, correlate zone diameter breakpoints could be selected to accurately distinguish susceptible wild-type strains from less-susceptible isolates. Examples of breakpoints for the 10-μg fusidic acid disk and the CLSI method of ≥21 mm for susceptibility and ≤18 mm for resistance (see solid vertical and horizontal lines in Fig. 1a) were published by Skov et al. (20). Applying these to the results in Fig. 1a, the absolute intermethod categorical agreement was 99.9%, with only one minor error. A slight adjustment to ≥22 mm (S) and ≤19 mm (R) produced complete (100.0%) intermethod accord. Using a higher susceptibility MIC of ≤1 μg/ml (13) and the same correlate zone diameters also yielded a very high level of intermethod agreement (99.7%), but the modification of the zone diameter criteria to ≥22 mm (S) and ≤19 mm (R) returned the agreement to 99.9%. These results are in close agreement with those criteria suggested by Toma and Barriault (21), also using the CLSI method, a 10-μg disk, and Mueller-Hinton medium.

Figures 1b and 2 demonstrate the excellent agreement for the 5-μg fusidic acid (CEM-102) disk results and reference broth microdilution tests (Fig. 1b) and the outstanding correlation between the 5- and 10-μg disk zone diameters (r = 0.97) (Fig. 2). Applying the breakpoint criteria suggested by Skov et al. (20) resulted in perfect (100.0%) intermethod agreement between the CLSI broth microdilution and the 5-μg disk results. Although the 5-μg disk for fusidic acid could certainly be standardized for use, the 10-μg disk is more widely available or preferred, with at least three manufacturers internationally.

Figure 3 shows the correlation of the fusidic acid (CEM-102) reference broth microdilution results with the MICs produced by Etest. The essential agreement was 99.7% ± one log₂ dilution step with 55.4% identical MIC results. A slight trend toward a lower MIC (31.2% of results were one log₂ dilution lower) was noted for the Etest. The Etest proved to be an acceptable alternative method to determine fusidic acid MIC results for S. aureus, with an intermethod agreement comparable to that for the CLSI disk diffusion method (e.g., >99%).

In summary, the in vitro diagnostic tests for fusidic acid (CEM-102) and S. aureus performed at an acceptable level of intermethod agreement. The CLSI M07-A8 (6) broth microdilution method performed well, as did the reference agar disk diffusion method of CLSI M02-A10 (5), each showing excellent intermethod categorical accuracy for either 5- or 10-μg disks. For the 10-μg disk, we propose zone diameter breakpoints of ≥22 mm (≤1 μg/ml), 20 to 21 mm (2 μg/ml), and ≤19 mm (≥4 μg/ml) for the susceptible, intermediate, and resistant category, respectively, which would provide harmonization with current EUCAST criteria (13) (Table 1). Alternatively, the interpretive criteria of Skov et al. (20) would provide a comparable level of accurate intermethod performance. The Etest could be applied as an alternative MIC method with near-complete concordance by quantitative measure (MIC) and by category analyses. The potency of fusidic acid can be assessed with confidence by the standardized CLSI MIC and disk diffusion test methods and by the Etest during clinical trials in the United States and elsewhere. Such testing will be important in monitoring emerging resistant subpopulations, such as those that have appeared in several nations over the last few decades (15).

• Open in new tab
• Download powerpoint

FIG. 1.

(a) Scattergram comparing fusidic acid (CEM-102) broth microdilution MIC results with zone diameters obtained with a 10-μg fusidic acid disk for 778 isolates of S. aureus. The solid lines indicate the interpretive breakpoints proposed by Skov et al. (20). The broken lines indicate alternative MIC (≤1 μg/ml) and disk diffusion (≥22 mm) interpretive criteria consistent with EUCAST guidelines (13). Values show the number of isolates with each result. (b) Scattergram comparing fusidic acid (CEM-102) broth microdilution MIC results with zone diameters obtained with a 5-μg fusidic acid disk for 728 isolates of S. aureus. The solid lines indicate the interpretive breakpoints proposed by Skov et al. (20), and broken lines indicate a proposed higher MIC breakpoint of ≤1 μg/ml. Values show the number of isolates with each result.

• Open in new tab
• Download powerpoint

FIG. 2.

Scattergram showing the excellent correlation obtained with a 5-μg and 10-μg fusidic acid disk diffusion tests (728 strains). Values show the number of isolates with each result.

• Open in new tab
• Download powerpoint

FIG. 3.

Comparison of fusidic acid broth microdilution and Etest MIC results for 728 isolates of S. aureus (r = 0.77). Values show the number of isolates with each result. Solid lines represent the ± 1 log₂ dilution values for equivalent or identical MICs.

• Copyright © 2010 American Society for Microbiology