Candida albicans isolates from 402 patients with no prior history of treatment with 5-fluorocytosine were collected at five medical centers from different areas of the United States. Isolates could be separated into four groups based on their minimum inhibitory concentrations (MICs) to 5-fluorocytosine. Group I isolates (60%) had MICs less than or equal to 12.5 micrograms/ml after 7 days, whereas groups II (22%), III (14%), and IV (4%) demonstrated MICs greater than 12.5 micrograms/ml on days 7, 2, and 1, respectively. Serotypes A and B accounted for 50.7 and 49.3%, respectively, of the 398 isolates typed. Serotype B was less prevalent in group I (26%), but predominated in the more resistant groups, groups II (85%), III (86%), and IV (53%). The common practice of identifying as "resistant" those isolates with MICs greater than 12.5 micrograms/ml after 48 h of incubation would yield a resistance rate in the United States of 11.5 to 15.5% in four centers and 35% in the fifth. Although serotype B and small agar disk diffusion zone sizes correlated with poor 5-fluorocytosine susceptibility, their ability to predict tube dilution MICs was limited. The true predictive value of such tests awaits correlation with in vivo studies.

The in vitro susceptibility of Candida albicans isolates to flucytosine was compared to therapeutic effect in experimental murine candidiasis (candidosis). Four groups of 10 isolates were chosen, based upon their broth dilution minimal inhibitory concentrations (MICs), from a group of 402 isolates from patients without prior flucytosine therapy. Group I MICs were less than 12.5 micrograms/ml after seven days, whereas group II, III, and IV MICs exceeded 12.5 micrograms/ml on days 7, 2, and 1, respectively. Pilot experiments selected challenge inocula of similar virulence. Mice were infected intravenously and given various flucytosine doses. Significant prolongation of survival correlated with MICs and with agar disk-diffusion zone diameters (P less than 0.05). In vivo response to therapy was more favorable for group I isolates compared with group IV isolates (P less than 0.01). The present study demonstrates in this animal model that in vitro susceptibility does correlate with in vivo response to therapy, although exceptions occur with individual isolates.

Strain phenotypes of 330 Candida albicans isolates from five areas in the United States were determined on the basis of nine biochemical tests. Statistical analysis of the distribution of phenotypes revealed no significant differences among types from different anatomical sources. However, there were some differences among the phenotypes of strains from the different geographical areas, and there were substantial differences in biochemical phenotypes associated with strains susceptible and resistant to 5-fluorocytosine and between strains of serotypes A and B. Geographical differences in phenotypes of C. albicans were also noted between the 330 U.S. isolates and 247 isolates from Britain. Cluster analysis of the U.S. strains alone and of all of the U.S. and U.K. strains showed that C. albicans phenotypes can be grouped into fewer than 20 clusters with common biochemical properties.

Sequential lung injuries, such as oxygen toxicity followed by septicemia, are common during the adult respiratory distress syndrome (ARDS). As these forms of vascular injury may be mediated in part by polymorphonuclear leukocytes (PMN), aberrant interactions between PMN and previously injured pulmonary endothelium are of both theoretical interest and clinical importance. The present study was undertaken to test the hypothesis that early oxygen toxicity at a dose that injuries pulmonary endothelium relatively selectively alters intrapulmonary neutrophil kinetics. Unanesthetized rats breathing 1.0 atmospheres oxygen for 36 h showed ultrastructural endothelial damage but no edema, injury, or neutrophilic inflammation by histologic criteria. However, in these oxygen-toxic animals, whereas initial accumulation of radiolabeled PMN in lungs was normal, washout of PMN was abnormal at 120 min after infusion, at which point the pulmonary retention of radiolabeled PMN in the lungs of oxygen-treated animals was significantly higher than in control animals (139% of control, p less than 0.0096). Features of our methodology, including avoidance of osmotic stress and use of paired control animals, appear to have greatly enhanced the sensitivity of radiolabeled neutrophils for detecting a subtle abnormality of neutrophil-endothelial interactions. Our studies in the oxygen toxicity model provide the first demonstration in vivo of abnormal intrapulmonary neutrophil kinetics in early oxygen toxicity prior to the onset of histologic evidence of lung injury or inflammation.

No abstract