Abstract
We assessed ceftaroline disk diffusion breakpoints for Staphylococcus aureus when applying revised Clinical and Laboratory Standards Institute (CLSI) ceftaroline MIC breakpoints. Disk-MIC correlation was evaluated by testing a challenge collection (n = 158) of methicillin-resistant S. aureus (MRSA) isolates composed of 106 randomly selected isolates plus 52 isolates with decreased susceptibility to ceftaroline (MIC, 1 to 16 μg/ml). Disk diffusion was performed with 30-μg disks and Mueller-Hinton agar from 2 manufacturers each. Revised CLSI susceptible (S)/susceptible dose-dependent (SDD)/resistant (R) MIC breakpoints of ≤1/2 to 4/≥8 μg/ml were applied. The disk breakpoints that provided the lowest error rates were CLSI S/R breakpoints of ≥25 mm/≤19 mm, with no very major (VM) or major (Ma) errors and with minor (Mi) error rates of 0.0% for ≥2 doubling dilutions above the I or SDD (≥I + 2), 22.1% for I or SDD plus or minus 1 doubling dilution (I ± 1), and 2.3% for ≤2 doubling dilutions below the I or SDD ≤I - 2 (overall Mi error rate, 16.5%). No mutation in the penicillin-binding protein 2a (PBP2a) was observed in 5 of 15 isolates with a ceftaroline MIC of 2 μg/ml; 3 of 11 isolates with a ceftaroline MIC of 1 μg/ml exhibited mutations in the penicillin-binding domain (PBD; 1 isolate) or in the non-PBD (2 isolates). All isolates except 1, with a ceftaroline MIC of ≥4 μg/ml, showed ≥1 mutation in the PBD and/or non-PBD. In summary, results from the disk diffusion method showed a good correlation with those from the reference broth microdilution method. Our results also showed that the ceftaroline MIC distribution of isolates with no mutations in the PBP2a goes up to 4 μg/ml, and reference broth microdilution and disk diffusion methods do not properly separate wild-type from non-wild-type isolates.