Iris K. Minichmayr (1), Lena E. Friberg (1)
(1) Department of Pharmacy, Uppsala University, Box 580, 75123 Uppsala, Sweden
Objectives: Pharmacokinetic-pharmacodynamic (PKPD) models based on in vitro time-kill studies of antibiotics have emerged as indispensable tools to identify efficacious therapeutic dosage regimens. Unlike traditional time-collapsed approaches of translation, relying on summary measures like PK/PD indices including fT>MIC (time that unbound concentrations exceed the minimum inhibitory concentration MIC), PKPD models enable to mimic the impact of an antibiotic regimen and the associated human concentration-time profile on the entire time course of bacterial growth and killing. This proves particularly useful if resource-intensive randomised controlled trials are unfeasible, as in the case of relatively rare drug-resistant pathogens or highly specific patient populations. Furthermore, simulations can aid in pre-selecting promising scenarios to be studied in clinical trials. This study sought to assess the expected relative benefit of increasingly popular prolonged-infusion dosing strategies for the beta-lactam meropenem (MEM) given diverse clinically relevant (patient- and pathogen-related) scenarios and populations. We thereby aimed to compare conclusions using a semi-mechanistic PKPD model-based versus a time-collapsed PK/PD index-based approach for translation to inform human antibiotic dosing.
Methods: We developed a showcase platform based on a PKPD model describing MEM effects against resistant Pseudomonas aeruginosa (ARU552, MIC=16 mg/L) over time (24 h). The model comprised two bacterial subpopulations, each with compartments for growing and resting bacteria, and had successfully been shown to predict animal data [1]. Bacterial growth and kill during MEM exposure were translated to humans, i.e. assumed to be driven by PK profiles of patients, based on six published population PK models [2-7]. A typical adult infected population, original PD parameters, and an initial bacterial load of 106 CFU/mL were chosen as a default scenario. Alternative scenarios comprised varied pathogen features (e.g. EC50, growth rates) and patient characteristics (e.g. creatinine clearance CLCR). For each scenario, total bacterial load (Btot) and fT>MIC were determined for ten dosage regimens (total daily doses TDD 3000-6000 mg; 0.5 h-, 3 h-, and continuous infusions±loading doses CI±LD) at 8 h and 24 h after start of dosing. Simulations (n=1000) were conducted considering between-patient-PK-variability and using R3.6.1 (mrgsolve package).
Results: The dosage regimen identified as most favourable (highest fT>MIC or lowest Btot) differed depending on (i) the measure of efficacy evaluated (fT>MIC or Btot), (ii) the time of assessment (24 h or 8 h) and (iii) the representative in the population assessed (median or P0.95, i.e. with 95% of the population meeting a target). Generally, 3-h infusions performed better (or comparably for TDD3000mg) than standard 0.5-h infusions. CI3000mg without LD appeared as the least favourable dosage regimen for the ARU552 strain. Given MIC=16 mg/L, CI6000mg+LD1000mg appeared as the most favourable regimen for the default scenario and the five other ‘typical’ populations, except when assessing fT>MIC_P0.95 (fT>MIC reached by 95% of patients), for which the 3-h6000mg regimen performed best for most scenarios (at both 24 h and 8 h). High TDD and shorter infusion durations (3-h6000mg) tended to become advantageous for scenarios associated with lower MEM concentrations (CLCR250mL/min, high CL), assessment of PI0.95 (Btot, fT>MIC), and lower bacterial susceptibility (MIC≥16 mg/L). In contrast, continuous infusion appeared to become more favourable for scenarios associated with higher MEM concentrations (CLCR30mL/min), assessment of the median fT>MIC and Btot, and higher bacterial susceptibility (MIC≤8 mg/L). For MIC≤8 mg/L, CI3000-6000mg regimens (MIC4mg/L) or CI6000mg regimens (MIC8mg/L) overall performed best. A discrepancy between conclusions based on fT>MIC and Btot was more pronounced at 8 h than at 24 h (particularly after a LD, foremost improving Btot at 8 h). As expected, all dosage regimens appeared more effective given 40% lower bacterial growth (and death) rates.
Conclusions: We present a showcase platform exemplified by meropenem and P. aeruginosa to illustrate how semi-mechanistic PKPD models can serve to translate in vitro antibiotic efficacy over time to clinical situations and to consequently identify the most promising strategies for efficacious dosing, also taking into account factors beyond PK.
References:
[1] Mohamed AF et al. J Antimicrob Chemother 71:1279-1290 (2016)
[2] Li et al. J Clin Pharmacol 46:1171-1178 (2006)
[3] Doh et al. J Antimicrob Chemother 65: 2428-2435 (2010)
[4] Delattre et al. Clin Biochem 45: 780-786 (2012)
[5] Lu et al. Antimicrob Agents Chemother 60: 6619-6625 (2016)
[6] Roberts et al. J Antimicrob Chemother 64: 142-150 (2009)
[7] Wittau et al. Antimicrob Agents Chemother 59: 6241-6247 (2015)
Reference: PAGE 30 (2022) Abstr 9964 [www.page-meeting.org/?abstract=9964]
Poster: Drug/Disease Modelling - Infection