As multidrug resistance in Gram-negative bacilli increases, colistin has rapidly gained much attention as one of few last line treatment options. An important part of therapeutic regimens with colistin has been the use of combinations, although there are no reliable methods available for testing combinations, and that the interpretation of joint effects is lacking. In addition, results of combination tests have been varying, with some results indicating no benefit of combinations while others indicate the opposite. The aims of this study were to describe the PD of colistin and meropenem in combination using clinically relevant concentrations of the drugs, to investigate some key endpoints for measuring effect, and to propose a possible path toward testing clinical strains with regard to joint effects.
Eight isolates of P. aeruginosa and A. baumannii, with and without meropenem resistance, were exposed to meropenem and colistin separately and in combination over one dosage interval in an in vitro kinetic model. The PD endpoints were initial kill rate, maximum kill, net kill and area under the bactericidal curve (AUBC). Checkerboard assays were performed for comparison.
Compared to the single drug exposures, the combination increased the effect with all endpoints in three of the strains, including one resistant to meropenem. The AUBC was significantly (paired t-test, p<0.05) lowered in three of the strains.
The chosen endpoints reflect different features of drug effect, and may be important to discern in some patient populations. Kill rate, maximum kill and net kill are point-based endpoints and focusing on only one of them (e.g. net kill by tradition) lead to wasting of potentially important information. AUBC reflects a compressed measure of the whole dosage interval that probably is relevant to most patients and can be used in evaluating joint effect easily and robustly by a simple t-test. The combination of meropenem and colistin is a useful option for treatment of multidrug resistant infections.