Clinical pharmacology of the new penicillins: I. The importance of serum protein binding in determining antimicrobial activity and concentration in serum

Publisher Website
Google Scholar
Abstract
The effect of human serum binding on free and total concentrations and urinary excretion of penicillins G, V, and ampicillin, oxacillin, cloxacillin, dicloxacillin, and methicillin achieved after administration in man is reported. In vitro studies comparing minimal inhibitory and bactericidal activity of these penicillins in 100 per cent human serum and trypticase soy broth revealed a close relation of inhibition of antimicrobial action to the extent of binding to serum. Quantitative data were obtained by use of arithmetic linear dilutions, in small steps, rather than the usual twofold dilution method. Extensive binding to serum proteins was demonstrated by equilibrium dialysis and ultrafiltration methods utilizing serum obtained from human volunteers receiving therapeutic doses of each penicillin analogue. Dicloxacillin, cloxacillin, oxacillin, and nafcillin were most highly bound in descending order. Penicillin G and V were intermediate, while methicillin and ampiCillin were least bound. Serum binding appears to be an important determinant of the high serum concentrations achieved with some of the new orally absorbed penicillinase‐resistant compounds. The uncritical use of the term “absorption curve” to describe concentrations of drugs which differ in distribution, metabolism, and excretion, after oral administration is deplored since the term implies that serum concentrations are the resultant of efficiency of gastrointestinal absorption alone. Alternate methods to assess absorption are illustrated. The customary twofold dilution broth assay of serum antibacterial activity may be misleading when employed with highly bound antibiotics. It is proposed that reports of studies of serum concentrations achieved with highly bound antimicrobial agents include data not only on total drug levels, but of free as well. This approach permits a more realistic correlation of serum concentration with minimum effective levels determined in the usual broth assays.