Use of Monte Carlo simulations to select therapeutic doses and provisional breakpoints of BAL9141

Abstract

BAL9141, a new antimicrobial agent belonging to the class of parenteral pyrrolidinone-3-ylidenemethyl cephalosporins, is active against most gram-positive microorganisms, including methicillin-resistant variants (methicillin-resistant Staphylococcus aureus [MRSA] and methicillin-resistant Staphylococcus epidermidis [MRSE]), as well as against penicillin-resistant pneumococci (PRP) and many gram-negative microorganisms. BAL9141 is administered as the prodrug BAL5788, which is rapidly converted to BAL9141 by plasma esterases. Pharmacokinetic (PK) data obtained in a previous multiple ascending dose study were used to fit a population PK model to using the NPEM2 program, yielding PK parameter estimates and its covariance matrix for BAL9141. These estimates and matrix were used to perform Monte Carlo simulations (MCSs) and obtain unbiased target attainment rates (TARs) for various time periods during which the concentration remains above the MIC (T(>MIC)). Assuming a T(>MIC) of 40%, TARs of 100% were reached with a dose of 500 mg/liter every 12 h for pathogens with MICs of 2 mg/liter and with a dose of 750 mg/liter every 12 h for pathogens with MICs of 4 mg/liter. Because MICs are </=2 mg/liter for most strains of MRSA, MRSE, and PRP (with some strains showing an MIC of 4 mg/liter), a dosing regimen of 750 mg every 12 h is proposed for clinical studies. The corresponding provisional breakpoint is S (susceptible) </= 4 mg/liter.

FIG. 1.

Plot of predicted versus observed concentrations (in nanograms per milliliter) of BAL9141 for 12 patients by the population model, based on experimental data. The r² value was 0.98 (P < 0.0001). The x axis shows the predicted concentrations based on parameter medians for the distribution of individual subjects. The figure shows the individual datum points, the regression line, and the y = x line for the entire population. The regression line superimposes on the y = x line.

FIG. 2.

T_>MIC as a function of MIC for various dosing regimens of BAL9141. Various dosing regimens of BAL9141 were used: 250, 500, 750, and 1,000 mg/liter every 8 or 12 h.

FIG. 3.

Simulation of a dosing regimen of 750 mg every 12 h (750 q12), with 95% CI based on SDs without the full covariance matrix and the 95% CI as measured from samples from subjects.

FIG. 4.

TARs and 95% CI (thus, attainment rates at 2.5 and 97.5%, respectively) as obtained with simulations using means and SDs with or without the full covariance matrix.