Human host defense and cytokines in mycobacterial infectious diseases: interleukin-18 cannot compensate for genetic defects in the interleukin-12 system.

q8h for various half-lives from 1.5 to 7.3 h, based on data from [9] and [10]. The A/T ratio varies from 0.5 h/% to a little more than 2 h/% for the AUC/MIC range 25–225 h and is relatively independent of the half-life of the drug. Thus, the results clearly show that there is no justification for using the AUC/MIC ratio to predict the effect of a drug for which is the dynamically time 1 MIC linked variable. The principal point and explanation is that a change in MIC will result in a change in both the AUC/MIC ratio and the ; however, the time 1 MIC AUC/MIC ratio changes linearly, whereas the changes log-linearly betime 1 MIC cause of the exponential nature of the elimination of most antibiotics. Thus, doubling the MIC will halve the AUC/ MIC ratio, but the change in the time 1 will depend on the half-life of the MIC drug. In general, if there is not too much accumulation of the drug, it will decrease by 1 half-life. The following additional notes can be made. (1) If the complete dose-effect relationships are used, rather than only those values for AUC/MIC ratios of 25–225 h, then the mismatch between the AUC/MIC ratio and the intime 1 MIC creases dramatically, with A/T ratios ranging from !0.5 to 1100 h/%. (2) The value of the constant (C) that is presumed to be equal to the A/T ratio is unknown. Depending on the value of this ratio, the “true” effect is either underestimated (i.e., the A/T ratio is !C) or overestimated (i.e., the A/T ratio is 1C). Because of the units used for AUC/MIC (i.e., hours) and (i.e., percent of time) the pretime 1 MIC sumed value of C is 11 in the range of interest. (3) The relationship between concentration and effect shows a steeper slope for b-lactam antibiotics than for antimicrobials with effects that are dependent on the AUC/MIC ratio; therefore, for blactam antibiotics, the change in the A/T ratio does not reflect a similar change in the probability of a positive response over the range of change of the AUC/MIC ratio. Thus, for an AUC/MIC ratio somewhere in the range of 25–225 h, a relatively small change in the will result time 1 MIC in an increase in effect that is relatively larger than might be expected from the change in the ratio itself. In conclusion, if the truly dynamically linked index for an antimicrobial is the , it makes no sense to use the time 1 MIC AUC/MIC ratio to predict the effect of either monotherapy or combination therapy. Either the true effect is underestimated or it is overestimated; in either case, the values of the AUC/MIC ratio will be misleading.