Raphaël Saporta (1), Elisabet I. Nielsen (1), Jon U. Hansen (2), Edgars Liepinsh (3), Iris K. Minichmayr (1), Lena E. Friberg (1)
(1) Department of Pharmacy, Uppsala University, Sweden, (2) Bacteria, Parasites & Fungi, Statens Serum Institut, Copenhagen, Denmark, (3) Latvian Institute of Organic Synthesis, Riga, Latvia
Objectives: Carbapenem resistance represents a serious clinical concern, with multiple carbapenem-resistant species listed as a critical priority for research by the WHO [1]. Although model-based approaches have been used to describe meropenem pharmacokinetics-pharmacodynamics (PKPD) in vivo [2-3], the time-course of the meropenem effects in vivo, notably against resistant strains, has not been characterized. This study aimed to develop a PKPD model describing the effect of meropenem over time on in vivo bacterial dynamics of meropenem-resistant strains. We further sought to use the developed PKPD model to predict PK/PD indices and expected efficacy of high-dose meropenem against the studied strains in patients.
Methods: Data were obtained from a neutropenic mouse thigh infection model. Meropenem plasma concentrations were measured in infected mice (n=30) over 2 hours after administration of a single dose of meropenem. In efficacy experiments, meropenem effects against 6 meropenem-resistant Escherichia coli or Klebsiella pneumoniae strains, with minimum inhibitory concentrations (MIC) ranging from 32 to 128 mg/L were evaluated. Mice (n=582) received meropenem doses ranging from 1 to 512 mg/kg as a single or repeated subcutaneous administration every 1 to 8 hours, or no treatment (growth control). Bacterial counts were obtained at a single time point per mouse, up to 24 hours after the beginning of treatment.
A PK model for meropenem in mice was built, and subsequently a PKPD model was developed to describe bacterial counts depending on meropenem concentrations. The PKPD model assumed that bacteria could be in a drug-susceptible and growing state (S) or a resting state (R) [3]. An initial delay in bacterial growth was tested by a steep increase of growth rate at an estimated time-point. A delay in drug effect was investigated through the addition of an effect compartment. Power and sigmoidal Emax models were evaluated to describe meropenem effects on bacterial killing. Differences in effect between strains were additionally evaluated using MIC values.
The final PKPD model was used to simulate dose fractionation studies in mice, and PK/PD indices were derived and correlated to 24-hour response. The PKPD model was used jointly with a PK model for meropenem in humans [5] to simulate PK/PD indices and outcomes after 24 hours of high-dose meropenem treatment with different infusion durations in normal/reduced kidney function patients (2g q8h/q12h).
Results: Meropenem PK was described by a 1-compartment PK model with linear elimination. The developed PKPD model included a growth delay as well as an effect compartment. The meropenem effect was described by an Emax model (Emax=0.67 h-1, EC50=1.01 mg/L) combined with an additional (sigmoid) Emax model, characterizing an apparent higher killing rate in the highest dose groups (Emax,2=2.04 h-1, EC50,2=88.8 mg/L, Hill=10). The MIC values measured in vitro could not be used to describe strain differences in meropenem effects. Visual predictive checks showed a good description of the data for all strains, dose groups and time-points.
In simulated dose fractionation studies in mice, the predictive ability of the fraction of time that unbound concentrations surpass a target concentration (fT>TC) was best for a concentration target equal to EC50, exhibiting higher correlations to response compared to MIC values of 32, 64, or 128 mg/L. For the simulations in humans, bacterial stasis or killing was predicted for most patients at 24 hours, with a trend towards less bacterial growth with prolonged durations of meropenem infusion compared to intermittent dosing.
Conclusions: The developed PKPD model successfully described the time-course of meropenem effects in vivo against strains with high MIC. Simulations illustrated limitations of the use of MIC as a PD marker, notably for PK/PD indices, when the in vitro MIC does not describe differences in in vivo susceptibility. Predictions in patients suggested an efficacy of high-dose meropenem monotherapy against the studied resistant strains.
References:
[1] Tacconelli E, Carrara E, Savoldi A, et al. Discovery, research, and development of new antibiotics: the WHO priority list of antibiotic-resistant bacteria and tuberculosis. Lancet Infect Dis 2018; 18: 318–27.
[2] Katsube T, Yamano Y, Yano Y. Pharmacokinetic-pharmacodynamic modeling and simulation for in vivo bactericidal effect in murine infection model. J Pharm Sci 2008; 97: 1606–14.
[3] Louie A, Liu W, VanGuilder M, et al. Combination treatment with meropenem plus levofloxacin is synergistic against Pseudomonas aeruginosa infection in a murine model of pneumonia. J Infect Dis 2015; 211: 1326–33.
[4] Nielsen EI, Viberg A, Löwdin E, Cars O, Karlsson MO, Sandström M. Semimechanistic Pharmacokinetic/Pharmacodynamic Model for Assessment of Activity of Antibacterial Agents from Time-Kill Curve Experiments. Antimicrob Agents Chemother 2007; 51: 128–36.
[5] Trang M, Griffith DC, Bhavnani SM, et al. Population Pharmacokinetics of Meropenem and Vaborbactam Based on Data from Noninfected Subjects and Infected Patients. Antimicrob Agents Chemother 2021; 65: e0260620.
Reference: PAGE 32 (2024) Abstr 10869 [www.page-meeting.org/?abstract=10869]
Poster: Drug/Disease Modelling - Infection