In a previous study with
flucytosine (5-FC) susceptibility of 40 Candida albicans isolates in vitro proved statistically correlated with response in systemic murine
candidiasis in vivo, although exceptions occurred with individual isolates. For the present analogous study with
ketoconazole, 58 C. albicans isolates were used of which 38 were from the 5-FC study and 20 were added to equalize the numbers of serotype A (
n = 30) and B (n = 28) and to make the range of susceptibility in vitro to
ketoconazole continuous and wide. The widest range of
ketoconazole susceptibility was noted for the minimal inhibitory concentrations on Kimmig and
Casitone agars (0.015-256 micrograms/ml) and disk zone diameters on YNB
agar (0-54 mm), whereas with disk tests on other media, the range of 50% inhibitory concentrations, relative inhibition factors and MICs on serum
agar remained narrow and/or showed strong ties. The Spearman's rank correlation between the in vitro activities determined with the various parameters showed wide variation consistent with p values from less than 0.001 to greater than 0.05. The serotype B isolates generally were more susceptible than the A isolates (p less than 0.02 for the majority of parameters). Evaluation of response in vivo was hampered by the low activity of
ketoconazole on the murine
infection with any of the isolates, the range of the ED50's being only 10- greater than 100 mg/kg. The serotype B
infections exhibited significantly better response (p less than 0.05) than the serotype A
infections. The overall correlation (Spearman's rank) of the susceptibility test results in vitro with the response in vivo was poor (p less than 0.05 for almost all parameters) suggesting very limited if any precise predictive values of the susceptibility tests in vitro with
ketoconazole against C. albicans. However, the narrow range of the ED50 suggests relatively little variation in the response of the different isolates in vivo and similarly small variation was also noted in some of the tests in vitro.