| Abstrakt: |
Typically, in vitro pharmacokinetic/pharmacodynamic (PK/PD) models, used to evaluate PK/PD endpoints, are carried out using broth as a culture medium, as well as for the determination of MIC. It is known that different culture mediums could affect the MIC values compared to broth. The objective of this study was to evaluate the effect of milk when it is used as medium culture on pharmacodynamic parameters of marbofloxacin (MFX) against coagulase negative staphylococci (CNS) isolated from goat mastitis, in an in vitro static-PK/PD model. Thirteen strains of CNS isolated from goat mastitis were included. Time-kill curves were conducted using goat milk or Mueller Hinton broth as culture medium in tubes supplemented with MFX in order to achieve different AUC24/MIC values, based on previous MIC values determined by microdilution method on Mueller Hinton broth: 0h (control), 3h, 12h, 24h, 48h, 96h and 192h. The area between the control growth and time-kill curves (IE) was used to quantify the antimicrobial effect of each AUC24/MIC. Then, AUC24/MIC vs. IE data was analysed by a nonlinear mixed-effect modelling with Monolix program (Lixoft, Batiment D, Antony, France). A Emax model with constant baseline and constant error model was selected. The Emax model is described below: E = S0 + ((Emax Cγ)/(Cγ C50γ)) Where S0 is the baseline effect; Emax is the maximal effect; C is AUC24/MIC value; C50 is the AUC24/MIC value that produces 50% of Emax and γ is a sigmoidicity factor. Culture medium was included in the analysis as a covariate, in order to study its effect on the model parameters. Covariate were retained in the final model if showed statistical significance (p<0.05), reduced the Inter Individual Variability (IIV) and reduced 2X log-likelihood, Akaike Information Criteria and Bayesian Information Criteria values. The results of the analysis showed that using milk as culture medium reduced C50 (drug potency) and γ, without significant effect on the other parameters, that is in concordance with higher MIC values reported in milk. Scatter plots of population/individual predicted vs. observed concentrations, population/individual weighted residuals vs. predictions/time, showed residuals little biased and uniformly distributed around and predictive value. This findings suggest that PK/PD endpoints should be higher compared with those determined in broth for mastitis produced by CNS. [ABSTRACT FROM AUTHOR] |
