Disk diffusion susceptibility testing of two macrolide antimicrobial agents: revised interpretive criteria for erythromycin and preliminary guidelines for roxithromycin (RU 965).

The 15-micrograms erythromycin disk was twice evaluated for interpretive accuracy against 417 and then 266 strains of gram-positive cocci, Neisseria meningitidis, and Haemophilus influenzae by using the criteria suggested by the National Committee for Clinical Laboratory Standards. These studies suggest a revision of streptococcal and Staphylococcus sp. interpretive guidelines to criteria (greater than or equal to 23 mm = susceptible, less than or equal to 13 mm = resistant) that are more compatible with in vivo erythromycin concentrations. It is also recommended that zone diameters be modified for H. influenzae (greater than or equal to 23 mm = susceptible, less than 22 mm = resistant) and that meningococci not be tested. A wide moderately susceptible category (1.0 to 4.0 micrograms/ml) would primarily include enterococcus strains and those organisms that would then respond only to parenteral administration of erythromycin. Roxithromycin (RU 965 or RU 28965), a new oxime ether erythromycin derivative, was also evaluated by investigator-prepared 15-micrograms disks and later with 30- and 60-micrograms commercial disks. Although roxithromycin was comparable to erythromycin in activity and regression line statistics, changes in the susceptible disk criteria were necessary because of superior roxithromycin serum concentrations and a longer serum half-life. Preliminary susceptible breakpoint criteria were greater than 21 mm = susceptible, 10 to 20 mm = indeterminate, and less than or equal to 9 mm = resistant. By using the recommended interpretive criteria for both macrolides, less than 98% absolute agreement was obtained, therefore suggesting the application of a spectrum class concept.