A Model-Based Pharmacokinetic/Pharmacodynamic Analysis of the Combination of Amoxicillin and Monophosphoryl Lipid A Against S. pneumoniae in Mice

Abstract

Combining amoxicillin with the immunostimulatory toll-like receptor 4 agonist monophosphoryl lipid A (MPLA) represents an innovative approach for enhancing antibacterial treatment success. Exploiting pharmacokinetic and pharmacodynamic data from an infection model of Streptococcus pneumoniae infected mice, we aimed to evaluate the preclinical exposure-response relationship of amoxicillin with MPLA coadministration and establish a link to survival. Antibiotic serum concentrations, bacterial numbers in lung and spleen and survival data of mice being untreated or treated with amoxicillin (four dose levels), MPLA, or their combination were analyzed by nonlinear mixed-effects modelling and time-to-event analysis using NONMEM^® to characterize these treatment regimens. On top of a pharmacokinetic interaction, regarding the pharmacodynamic effects the combined treatment was superior to both monotherapies: The amoxicillin efficacy at highest dose was increased by a bacterial reduction of 1.74 log10 CFU/lung after 36 h and survival was increased 1.35-fold to 90.3% after 14 days both compared to amoxicillin alone. The developed pharmacometric pharmacokinetic/pharmacodynamic disease-treatment-survival models provided quantitative insights into a novel treatment option against pneumonia revealing a pharmacokinetic interaction and enhanced activity of amoxicillin and the immune system stimulator MPLA in combination. Further development of this drug combination flanked with pharmacometrics towards the clinical setting seems promising.

Keywords: MPLA; amoxicillin; immunomodulation; murine model; pharmacometric PK/PD modelling; time-to-event modelling.

Figure 1

Schematic illustration of the pharmacometric nonlinear mixed-effects pharmacokinetic/pharmacodynamic model for amoxicillin (AMX) with coadministration of monophosphoryl lipid A (MPLA) of mice infected with Streptococcus pneumoniae serotype 1, comprising a two-compartments pharmacokinetic (PK) model (left), an effect compartment as PK/pharmacodynamic (PD) link model (middle) and a disease and treatment model including unrelated killing and natural death effects including killing effects of the immune system, and killing effects of AMX and MPLA (right). Abbreviations: Bacteria_lung: Number of bacteria in lung; Bacteria_spleen: Number of bacteria in spleen; C_e,lung: AMX concentration in lung effect compartment; C_e,spleen: AMX concentration in spleen effect compartment; CL_AMX: Clearance of AMX; Dose: Administered dose of AMX by oral gavage; EC₅₀: Concentration of AMX to achieve half maximum killing effect; E_max: Maximum killing effect of AMX; FC_{AMX + MPLA}: Fractional change of CL_AMX in presence of MPLA depending on the AMX dose implemented as covariate; Gut: Organ of AMX administration; H_lung: Hill factor for lung; H_spleen: Hill factor for spleen; k_a: First-order absorption rate constant; k_AMX: First-order killing rate constant of AMX in spleen representing the slope of the effect compartment concentration and effect relationship; k_e0,lung: First-order rate constant for effect delay in lung; k_e0,spleen: First-order rate constant for effect delay in spleen; k_g: First-order growth rate constant in lung; k_kill,lung: First-order rate constant for treatment-unrelated killing and natural death effects in lung; k_lag: First-order rate constant for delay in onset of bacterial growth in lung; k_MPLA,spleen: First-order killing rate constant for killing effect in spleen only in presence of MPLA; MPLA_lung: Fractional change of k_kill,lung in presence of MPLA; n: Number of transit compartments; Q: Intercompartmental clearance; t_lag: Lag time; V_c: Central volume of distribution; V_p: Peripheral volume of distribution.

Figure 2

Visual predictive check (n = 1000 simulations including unexplained variability) of the pharmacometric pharmacokinetic/pharmacodynamic model for bacterial numbers in lung (A) and spleen (B) stratified into study groups including the fractions of samples being below the LLOQ for lung (C) and spleen (D). Circles: Observations; Lines: 50th percentile (solid), 5th and 95th percentile (dashed) of observed (red) and simulated (black) bacterial numbers. 90% confidence interval around simulated percentiles as shaded area. Abbreviations: AMX: Amoxicillin (0.40 mg/kg or 1.20 mg/kg); LLOQ: Lower limit of quantification; MPLA: Monophosphoryl lipid A (2.00 mg/kg); +: Treatment with respective drug; –: No treatment with respective drug.

Figure 3

Visual predictive check of the overall survival model (n = 1000 simulations including unexplained variability) for different study groups of mice infected with Streptococcus pneumoniae serotype 1 and untreated or treated with AMX with or without coadministration of MPLA: Solid lines: Observed survival; Dashed lines: Simulated survival; 90% confidence interval around simulated survival as shaded area. Abbreviations: AMX: Amoxicillin (0.40 mg/kg or 1.20 mg/kg); MPLA: Monophosphoryl lipid A (2.00 mg/kg); +: Treatment with respective drug; –: No treatment with respective drug.