Introduction Neisseria gonorrhoeae shows increasing resistance to first line empirical treatment, which demonstrates the need for robust methods to evaluate antibiotic treatment regimens. Antibiotic efficacy is traditionally determined in vitro by measuring minimum inhibitory concentrations (MICs). Time-kill curve assays for N. gonorrhoeae have been difficult to standardise. We developed a new time-kill curve assay and used pharmacodynamics functions to analyse the relationship between antibiotic concentration and bacterial net growth rate.

Methods We used a defined medium (Graver-Wade medium) and grew bacteria in 96-well microtiter plates. To measure colony forming units over a time course of six hours, we used a previously described drop plate method and spotted 10 µl droplets on chocolate agar with a multichannel pipette. The assay was validated using 16 reference strains, representing diffent clades of the N. gonorrhoeae phylogenetic tree. We then studied a highly sensitive strain isolated in 1954 in Denmark in detail. Penicillin G, spectinomycin, gentamicin, tetracyline, chloramphenicol, ciprofloxacin, azithromycin, cefixime and ceftriaxone were examined in concentrations from 0.016× to 16×MIC. A pharmacodynamic function was fitted to the net bacterial growth rates at each concentration, resulting in four parameters that describe the pharmacodynamic properties of each antibiotic.

Results Our time-kill curve assay was reproducible for all the N. gonorrhoeae strains tested. Ciprofloxacin and spectinomycin induced the strongest bactericidal effect during the first six hours. Tetracycline and chloramphenicol were the only antibiotics that showed a purely bacteriostatic effect. Differences in the shape of the pharmacodynamic functions illustrate the time and concentration dependent properties of the antibiotics.

Conclusion We developed a standardised, robust and quality-assured time-kill curve assay for N. gonorrhoeae. Pharmacodynamic functions allowed classification of nine antibiotics according to their antimicrobial properties. These methods can be used to investigate new antibiotic regimens and help to improve dosing strategies.

Disclosure of interest statement This work was supported by SystemsX, Switzerland, Örebro County Council Research Committee and the Foundation for Medical Research at Örebro University Hospital, Sweden.