Bacterial pathogens of interest to Defence are intrinsically highly virulent and can be difficult to treat. Assessment of a therapeutics’ ability to treat a biothreat agent infection has traditionally relied on in vitro data and animal models, as human clinical trials are not feasible. However, treatment regimens derived from animal models may not translate well into humans due to differences in their antibiotic pharmacokinetics. The in vitro Hollow Fibre Infection Model (HFIM) is able to bridge this gap, as it can precisely mimic human antibiotic pharmacokinetic parameters to assess a whole treatment regimens’ effect on its target bacterial infection. Consequently, the HFIM has a high correlation in being able to predict clinical outcomes, generating data that is accepted by both the FDA and EMA.
Melioidosis is caused by Burkholderia pseudomallei, an environmental pathogen endemic to Northern Australia and Southeast Asia. It is innately resistant to numerous antibiotics and can only be treated with ‘antibiotics of last resort’. Here we assessed standard clinical regimens of levofloxacin and piperacillin/tazobactam (TZP) to treat a virulent B. pseudomallei infection in the HFIM. Ultimately the q24d 750mg levofloxacin regimen failed, resulting substantial bacterial regrowth and levofloxacin resistance. In contrast a q8d 4000/500mg TZP regimen was able to suppress bacterial growth without generating resistance. HFIM dose escalation and fractionation studies using Burkholderia thailandensis, an avirulent close relative to B. pseudomallei, determined the PK/PD driver and target value. This approach identified an optimal human dosing regimen of q6d 3000/375mg TZP delivered over a 2.5h iv infusion to be a potentially new melioidosis treatment. Interestingly, TZP has been long theorized as a potential melioidosis treatment, but there is no published information about its efficacy in animal models or human clinical trials. Here lies the strength of the HFIM, in being able to rapidly assess prescribed human dose regimens against a range of biothreat agents to generate efficacy data predictive of clinical outcomes.