Development of in vitro susceptibility testing methods for gemifloxacin (formerly LB20304a or SB-265805), an investigational fluoronaphthyridone

Potent investigational fluoroquinolones require convenient, but accurate, d iagnostic tests for initially applied clinical trials. For this purpose, ge mifloxacin (formerly SB-265805, LB20304a) was tested by the reference dilut ion tests and standardized disk diffusion methods of the National Committee for Clinical Laboratory Standards (NCCLS) to establish interpretive criter ia. For rapid-growing pathogens, 986 organisms were tested by broth microdi lution MIG, and 5- and 10-mu g disk diffusion tests. Correlation (r) betwee n 5- and 10-mu g disk zone diameters was 0.99 (y = -0.22 to 0.99x) and the preferred 5-mu g disk zone/MIC scattergram produced a regression of y = 14. 8 to 0.41x (r = 0.93). At potential pharmacodynamics (C-max = 1.3 mu g/mL f or 320 mg dose) validated breakpoints Of less than or equal to 0.5 mu g/mL for susceptible and greater than or equal to 2 mu g/mL for resistant, corre late zones of greater than or equal to 17 mm and less than or equal to 13 m m produced rare serious interpretive errors (0.1%) and 96.7% absolute categ orical agreement. For 304 Streptococcus pneumoniae and 305 strains of other streptococci, the same breakpoints produced 100 and 99.1% categorical accu racy even when testing levofloxacin-resistant (MIC, greater than or equal t o 4 mu g/mL) strains. Interpretive breakpoints were proposed for Hemophilus influenzae (300 strains tested), with complete correlation between tests. Etest (AB BIODISK, Solna, Sweden) was compared in all experiments with the fastidious species and showed a trend toward higher values (twofold). Gemif loxacin in vitro susceptibility test methods seem to be accurate and with v ery acceptable intermethod agreement, supported by previously reported func tional quality control guidelines. (C) 1999 Elsevier Science Inc.