In vitro and in vivo efficacy of drugs against Bacteroides fragilis

s of dutch ph.d, theses In vitro and in vivo efficacy of drugs against Bacteroides fragilis B.A.C. DIJKMANS,* University of Leiden, September 27, 1984. Promoter: Prof. Dr. R. van Furth (Leiden), Co-promoter: Dr. H. Mattie (Leiden). Referees: Prof. Dr. M.F. Michel (Rotterdam) and Prof. Dr. G.P. van Rees (Leiden). With respect to the incidence of anaerobic microorganisms Bacteroides fragilis is the most important pathogen in human pathology. Although most clinically important anaerobes are sensitive to benzylpenicillin, B. fragUis is not. Drugs to treat B. fragilis infections have been developed only rather recently, and thus clinical experience is still limited. Quantitative comparison of the antibacterial efficacy of these drugs is mainly done on the basis of a few in vitro parameters, of which the minimal inhibitory concentration (MIC) is one. For several reasons a better in vitro parameter of antibacterial efficacy, for comparison with the in vivo situation, is based on short-term growth. Growth of untreated B. fragilis in vitro is exponential and therefore can be characterized by a single parameter as a function of time. However, in the presence of antibacterial drugs the growth rate is time dependent. Therefore the effect of a given concentration of an antibacterial drug was expressed by the following equation: log ratio = i x (t + e'I). In this equation the log ratio is the difference between the logarithm of the number of bacteria in the control culture and that in the presence of the drug, and i is a concentration dependent parameter: The value of i may be regarded as the final slope of the log ratio curve, or the difference between the growth rate in the control and the final growth rate in the presence of antibiotics, which may be negative in the case of bactericidal drugs. The results of the in vitro studies with various drugs are shown in Figure I. The relationship between the MIC and the short-term growth inhibition is different for antibacterial drugs with different modes of action, e.g. for metronidazole and clindamycin, and similar for drugs with the same mode of action, e.g. metronidazole and tinidazole (Fig. 1). The implication of these findings is that the relative potency of two drugs is only constant for drugs with similar modes of action, while for others it depends on the effect level. For bacteriostatic drugs like clindamycin the MIC gives a deceptively favourable impression in comparison with for instance metronidazole. These findings led to the conclusion that it would be impossible to predict relative efficacy in vivo of unrelated drugs from in vitro data. Therefore an experimental thigh muscle infection in mice was adapted for the study of B. fragilis infections. It turned out that growth of B. fragilis only occurred if i )< I O 8 to 2 )< I O n bacteria were injected into the thigh together with 1 x IO 6 to 2 x 1o 6 Escherichia coli, an aerobic micro-organism. After an initial decline in numbers rapid outgrowth of B. fragilis occurred after about I8 h. Drugs were administered 6 h after infection and the animals were sacrificed t8 h later to count bacteria in the homogenized muscle. These bacterial numbers were the basis for the calculation, according to a standard parallel-line bioassay procedure, of the relative potency of all othei drugs with metronidazole as a standard. The potency ratios thus calculated are given in Table I. Metronidazole and tinidazole were equally effective in vivo, although in vitro metronidazole was twice as effective as tinidazole (Fig. I ) . This might be explained by the longer half-life of tinidazole, leading to a slightly larger area under the plasma curve (AUC) of the concentrations of the nonprotein bound drug. Clindamycin was, on a dosage base, almost eight times less potent than metronidazole. However, when the large pharmacokinetic differences between the two drugs are taken into account, the activity of clindamycin in vivo proves to be only about 50% higher than that of metronidazole (Table 1). Apparently the limited maximal effect in vitro (Fig. 1) is more important than the high degree of activity of clindamycin at low effect levels, as expressed by the MIC. A similar but greater discrepancy was found for rifamycin SV. When the pharmacokinetic divergence, which favours rifamycin SV, is taken into account, rifamycin SV is almost forty times less potent than metronidazole (Table I). Again this is contrary to what would be expected by comparing MICs in vitro, but it corresponds well with the finding of the very low maximal effect in vitro (Fig. I). Vancomycin, although having a different mode of action, was quantitatively more comparable with *Leiden University Hospital, P.O. Box 9600, 2300 RC Leiden, The Netherlands. 170 Vol. 7 I985 Pharmaceutisch Weekblad Scientific Edition