Susceptibility testing of macrolide antibiotics against Haemophilus influenzae and correlation of in vitro results with in vivo efficacy in a mouse septicemia model.

There is poor correlation between the MICs and zone sizes obtained for erythromycin against Haemophilus influenzae. The effect of two media, Mueller-Hinton medium supplemented with 3% lysed horse blood and 10 micrograms of NAD per ml (MHA + LYHB) and Mueller-Hinton agar supplemented with 1% bovine hemoglobin and 1% IsoVitaleX (MHA + HGB), on the MICs and zone sizes of erythromycin against H. influenzae was determined. The effect of three different methods for inoculum preparation on the susceptibility of H. influenzae was also determined. The MICs were independent of the method of inoculum preparation, but the zone sizes were smaller if the inoculum was carefully adjusted to contain approximately 10(8) CFU/ml. MICs were higher and zone sizes were smaller when MHA + HGB was used instead of MHA + LYHB. Good correlation was found when MHA + LYHB was used for determining the MIC and MHA + HGB was used for determining susceptibility by the disk method. When the inoculum was adjusted to match a McFarland 0.5 standard, the viable counts had to be approximately 10(8) CFU/ml for good correlation between MICs and zone sizes. A-56268, a new macrolide antibiotic, was tested against H. influenzae, and its MICs and tentative breakpoints against this organism were determined. The MICs obtained by various methods were correlated with in vivo efficacy by using a mouse septicemia model. MICs obtained on MHA + HGB or MHA + LYHB incubated without a 5% CO2 atmosphere showed the best correlation with in vivo efficacy.