Chenyan Zhao (1), Pikkei Wistrand-Yuen (2), Pernilla Lagerbäck (2), Thomas Tängdén (2), Elisabet I. Nielsen (1), and Lena E. Friberg (1)
(1) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden, (2) Department of Medical Sciences, Section of Infectious Diseases, Uppsala University, Uppsala, Sweden
Objectives: The increasing prevalence of refractory multidrug resistant bacteria calls for new approaches to suggest effective antibiotic combination therapies. The combined effect of minocycline (MIN) plus polymyxin B (PMB) was identified to be promising in our previous screening procedure [1] and was thus selected for in vitro static time-kill studies (TKS) to further evaluate the effectiveness. To characterize the observed TKS data and quantify the drug interaction effect, we aimed to develop an in silico semi-mechanistic pharmacokinetic–pharmacodynamic (PKPD) model. Through simulations, the model was applied to explore the clinical potential of such drug combination to overcome highly resistant bacteria.
Methods: The clinical isolate Klebsiella pneumoniae ARU613, an extended-spectrum β-lactamase and carbapenamase producing strain, resistant to both MIN and PMB, was selected for the study. Minimum inhibitory concentrations (MICs) for MIN and PMB were 12mg/L and 16mg/L, respectively. ARU613 was first exposed to a wide range of concentrations (0–64 xMIC) of either MIN or PMB alone with the number of colony forming units per mL (CFU/mL) counted at pre-determined time points between 0-28 hours. The generated data was used for the development of single drug PKPD models, which after their effects were combined, guided the selection of a limited number of informative drug combination TKS to be performed. Thereafter, the drug combination TKS data were used to update the PKPD model and interaction functions (the power interaction model [2], Bliss independence model [3] and general PD interaction (GPDI) model [4]) were explored. L2 [5] and M3 [6] methods were used to handle replicate CFU counts and below limit of detection data, respectively. Modelling was conducted in NONMEM 7.4.2. Visual predictive checks (VPC) implemented in PsN 4.7.15 were generated for model evaluation. Reported clinical population PK models for MIN [7] and PMB [8] were subsequently connected to the final PKPD model to predict the combination drug effect in simulated patients with an initial bacterial load of 6.8 log10 CFU/mL.
Results: Both MIN and PMB single drug models were based on the self-limiting bacterial growth model [9] in which bacteria transfer between 2 compartments residing 1) drug-susceptible, growing bacteria and 2) non-drug-susceptible, non-growing bacteria. The observed regrowth at later time points was here described by adaptive resistance models for both drugs. A model with a resistant subpopulation did not describe the data equally well. The observed TKS data indicated a benefit for combining MIN and PMB against ARU613. A synergistic effect of MIN and PMB was identified in the PKPD model where the GPDI interaction model described the data the best; PMB enhanced the MIN bacterial killing effect in a concentration-independent manner while the impact of MIN on PMB could be ignored. The estimated interaction parameter, which inclusion reduced OFV by 62 units (P < 0.001, df = 1), indicated that MIN potency for bacterial killing increased 53.5% in the presence of PMB. VPC plots demonstrated that the developed model could adequately describe the observed CFU counts. Predictions based on human PK indicated that clinically available high dosage regimens of MIN+PMB could keep bacterial counts below the start inoculum for more than 24 hours for the investigated strain resistant to the individual antibiotics.
Conclusions: We successfully developed a semi-mechanistic PKPD model describing MIN and PMB interaction on carbapenem-resistant Klebsiella pneumoniae. Model predictions where concentrations from population PK models drove the bacterial killing supported clinical use of MIN and PMB in combination to overcome infections caused by highly resistant strains.
References:
[1] Yuen P et al ECCMID (2017) Abstr 161096 [http://m.eccmidlive.org/webapps/scientific.html#Abstracts/161096]
[2] Mohamed AF. J Antimicrob Chemother (2016) 71, 1279–1290
[3] Bliss CI. Ann Appl Biol (1939) 26, 585–615
[4] Wicha SG et al. Nat Commun. (2017) 8, 2129 [5] Karlsson MO et al. J PK Biopharm (1995) 23, 651–72
[6] Beal SL. J PKPD (2001) 28, 481–504
[7] Yamamoto T et al. Am J Ther (1999) 3, 157-60.
[8] Miglis C et al. Antimicrob Agents Chemother (2018), 62:3 12 e01475-17
[9] Nielsen EI et al. Antimicrob Agents Chemother. (2007) 51, 128-36
Reference: PAGE 27 (2018) Abstr 8745 [www.page-meeting.org/?abstract=8745]
Poster: Drug/Disease Modelling - Infection