We thank De Pascale et al for their interest in our article in CHEST (December 2010).1 Our study (and those by other authors) showed that after IV administration of colistin methanesulfonate (CMS), the plasma concentration of colistin is apparently suboptimal. The adverb “apparently” is necessary because the values of the pharmacodynamic indices that best predict the efficacy of (free) colistin have been determined in in vitro and animal studies,2,3 but are not yet known for humans. The results of in vitro pharmacodynamic studies of antibiotics, although very useful for correlating drug concentrations and pharmacologic effects, might not be directly translatable to the clinical setting. In the case of colistin, the picture is even more complex. In fact, colistin (like polymyxin B) binds to the lipopolysaccharides (LPS) released by killed bacteria also at concentrations below the minimum inhibitory concentration and which, therefore, have no effect on bacterial counts.4 This colistin-LPS binding could inhibit the toxic effects of the bacterial products, or contrariwise, could reduce the amount of available colistin and its bactericidal activity. The balance between the beneficial and harmful effects of LPS binding has not yet been defined.