Humoral immunity does not play a prominent role during experimental cryptococcosis. However, previous studies have shown that immunoglobulin G (IgG) anti-Cryptococcus neoformans antibodies can mediate cell-dependent yeast killing in vitro. Therefore, the protective effect of a previously described monoclonal IgGl anti-C. neoformans antibody (El) administered intraperitoneally 24 h before intravenous infection with a C.

Two monoclonal immunoglobulin G1 antibodies reacting with Cryptococcus neoformans capsular polysaccharide (CNPS) were produced in mice by using a carefully defined procedure for immunization with unmodified CNPS purified from C. neoformans serotype A. Since the antibodies were found to have the same pattern of specificity, only one of them (E1) is described. This anti-CNPS monoclonal antibody reacted with the glucuronoxylomannan component of CNPS but not with the constituent monosaccharides or with the mannose alpha(1----3)-linked oligosaccharide structures present on CNPS. E1 appeared to be specific for C. neoformans serotype A by agglutination of whole cells; it was specific for soluble CNPS A by gel immunoprecipitation. However, indirect immunofluorescence and competitive-binding enzyme-linked immunosorbent assay experiments showed low levels of cross-reactivity with serotypes B and D but not with serotype C. Concentrations 10,000 times higher for serotypes B and D cells than for serotype A cells were required for a 50% inhibition of E1 anti-CNPS A activity as measured by enzyme-linked immunosorbent assay. Among the other yeasts tested, a cross-reaction was only detected with Trichosporon beigelii. The four serotypes of C. neoformans could be distinguished based on intensities and patterns of fluorescence in an indirect immunofluorescence assay using the monoclonal anti-CNPS A antibody. Monoclonal anti-CNPS A antibodies could be useful for fundamental studies on the glucuronoxylomannan structure, as well as for clinical applications such as serotyping and possibly the serological diagnosis of cryptococcosis.

In a rabbit model of Escherichia coli endocarditis, we studied the penetration into infected vegetations and the antibacterial effect of ceftriaxone. Ceftriaxone was given at different dosages, alone or with an interfering agent, diclofenac, a nonsteroidal anti-inflammatory drug, to determine the predictive value of the antibiotic levels in serum or infected vegetations on the antibacterial efficacy. Diclofenac increased the serum terminal half-life of ceftriaxone and increased its extravascular diffusion in tissue cage fluid, as well as in infected vegetations, allowing us to obtain various antibiotic concentrations in the infected site. Two hours after the fourth injection, around the time of peak level in serum, we observed a linear relationship between (i) serum and local antibiotic levels in vegetations, (ii) In severe infections, a bactericidal effect at the site of infection is essential for effective cure, as demonstrated by Scheld and Sande in experimental pneumococcal meningitis in rabbits (23) and by Decazes et al. in Escherichia coli meningitis (8). In the latter study, ceftriaxone levels in cerebrospinal fluid had to be at least 10 to 20 times higher than its MBC to obtain efficacy. The in vivo bactericidal effect depends on in vitro antibiotic activity and on the pharmacokinetic properties of the antibiotic molecule, which include the degree of diffusion at the infected site and consequently the local antibiotic concentration. In a previous study comparing the efficacy of cefotiam, cefmenoxime, and ceftriaxone in experimental E. coli endocarditis, we have shown that the antibiotic local level/MBC ratio roughly correlated with the antibacterial effect and could represent an adequate basis to explain the differences observed among these three drugs in vivo (19). We have also demonstrated that ceftriaxone was the most efficient of these three cephalosporins, even when given at relatively long dosing intervals, i.e., 24 h. Apart from the demonstration of this threshold of activity, the precise relationship between the range of concentrations and the antibacterial effect has not yet been determined.The purpose of this work was to assess the predictive value of serum antibiotic levels for the antibiotic concentrations in vegetations and for the local antibacterial effect. At the same time, we attempted to establish a correlation between the ceftriaxone levels in infected vegetations and * Corresponding author. the antibacterial effect. Because it is impossible to study these principles in humans, we used the experimental model of E. coli endocarditis that we have previously described (7). Although this type of infection is uncommon in humans, the model is reliable and the infection is easily reproducible. As it provides a rigourous test of in vivo antibiotic efficacy on a severe infection due to a gram-negative bacillus, the model appeared to be appropriate for studying the properties of ceftriaxone, a long-acting broad-spectrum cephalosporin. We administered different doses of ceftriaxone alone or com...