Susceptibility patterns for amoxicillin/clavulanate tests mimicking the licensed formulations and pharmacokinetic relationships: do the MIC obtained with 2:1 ratio testing accurately reflect activity against β-lactamase-producing strains of Haemophilus i

Amoxicillin/clavulanate has recently undergone formulation changes (XR and ES-600®) that represent 14:1 and 16:1 ratios of amoxicillin/clavulanate. These ratios greatly differ from the 2:1 ratio used in initial formulations and in vitro susceptibility testing. The objective of this study was to determine if the reference method using a 2:1 ratio accurately reflects the susceptibility to the various clinically used amoxicillin/clavulanate formulations and their respective serum concentration ratios. A collection of 330 Haemophilus influenzae strains (300 β-lactamase-positive and 30 β-lactamase-negative) and 40 Moraxella catarrhalis strains (30 β-lactamase-positive and 10 β-lactamase-negative) were tested by the broth microdilution method against eight amoxicillin/clavulanate combinations (4:1, 5:1, 7:1, 9:1, 14:1, and 16:1 ratios; 0.5 and 2 μg/mL fixed clavulanate concentrations) and the minimum inhibitory concentration (MIC) results were compared with those obtained with the reference 2:1 ratio testing. For the β-lactamase-negative strains of both genera, there was no demonstrable change in the MIC values obtained for all ratios analyzed (2:1 to 16:1). For the β-lactamase-positive strains of H. influenzae and M. catarrhalis, at ratios ≥4:1 there was a shift in the central tendency of the MIC scatterplot compared with the results of testing 2:1 ratio. As a result, there was a 2-fold dilution increase in the MIC50 and MIC90 values, most evident for H. influenzae and BRO-1-producing M. catarrhalis strains. For β-lactamase-positive strains of H. influenzae, the shift resulted in a change in the interpretive result for 3 isolates (1.0%) from susceptible using the reference method (2:1 ratio) to resistant (8/4 μg/mL; very major error) at the 16:1 ratio. In addition, the number of isolates with MIC values at or 1 dilution lower than the breakpoint (4/2 μg/mL) increased from 5% at 2:1 ratio to 32-33% for ratios 14:1 and 16:1. Our results indicate that, for the β-lactamase-positive strains of H. influenzae and M. catarrhalis, the results of the amoxicillin/clavulanate reference 2:1 ratio testing do not accurately represent all the currently licensed formulations. Pharmacokinetic/pharmacodynamic (PK/PD) target attainment might be compromised when higher amoxicillin/clavulanate ratios are used clinically. With a better understanding of PK/PD parameters, reevaluation of the amoxicillin/clavulanate in vitro susceptibility testing should be considered by the standardizing authorities to reflect the licensed formulations and accurately predict clinical outcomes.