the dose that would be calculated if it was based upon a well-controlled PK study
where the %CV was 15%. The results of this exercise are shown in Fig.1b.This
Figure shows that, within the range of variances used in the simulated dataset, the
actual dose needed to achieve the targeted exposure was almost 2.5-times greater
when the CV was 100% as compared to the dose needed when the %CV was 15.
4.2 Mutation Selection Window
The in vivo relevance of a mutation selection window (MSW) hypothesis was first
presented by Baquero ( 1990 ), when he proposed the existence of a range of drug
concentrations within which organisms which had reduced susceptibility had a
survival advantage. In their review, Epstein et al. ( 2004 ) discuss how subsequent
studies by Baquero and Negri ( 1997 ), Zhao and Drlica ( 2001 ), and Dong et al.
( 1999 , 2000 ) served to define the boundaries for a “dangerous zone”, and they
described the relationship between mutational events and resistance to quinolones.
Similarly, in a rabbit pneumonia model, gatifloxacin was found to pose the greatest
risk of selecting for resistant strains when dosed in accordance with a regimen that
produced the longest duration of exposure within the MSW (Croisier et al. 2004 ).
Drlica and Zhao ( 2007 ) defined the MSW from three discrete concentrations:
l The MIC of the wild-type bacteria.
l Concentrations above the MIC of the wild-type bacteria, where there is a plateau
in killing due to the survival of the least susceptible microbial subpopulation.
l Concentrations at which even the least susceptible organisms are killed. The
latter is termed the mutation prevention concentration (MPC). At the MPC, a cell
must acquire two concurrent resistance mutations for selection of resistant
strains. Such an event is highly unlikely. The MPC is estimated as the drug
concentration that blocks growth when 10^10 cells are applied to agar.
The window hypothesis is important because dosing guided by traditional PK/PD
standards often places antimicrobial concentrations inside the MSW, where they
can selectively enrich resistant mutant subpopulations. This can adversely affect
drug response not only in the infected individual but also in subsequent hosts. Once
amplified, the mutant cells can disseminate to a fresh host. Consequently, bacterial
population expansion occurs and a new round of antimicrobial pressure can further
enrich the mutant population (Croisier et al. 2004 ). Ultimately, such a scenario can
lead to a predominance of the mutant subpopulation and loss of antimicrobial
effectiveness over time (Epstein et al. 2004 ). However, Zhao and Drlica ( 2008 )
emphasised that there is a distinction between absolute clinical resistance, where
the pathogen cannot be killed by the drug, versus resistant mutants, where the
pathogen may be killed at higher drug concentrations. Therefore, they suggested
that the indices of AUC/MPC or T>MPC may be preferable to those based upon
MIC values. It might be noted that there is not a clear relationship between MIC and
MPC values within and between bacterial species (Drlica et al. 2006 ). In fact, some
244 M. Martinez and P. Silley