Use of Monte Carlo Simulations to Determine Optimal Carbapenem Dosing in Critically Ill Patients Receiving Prolonged Intermittent Renal Replacement Therapy.

Pharmacokinetic/pharmacodynamic analyses with Monte Carlo simulations (MCSs) can be used to integrate prior information on model parameters into a new renal replacement therapy (RRT) to develop optimal drug dosing when pharmacokinetic trials are not feasible. This study used MCSs to determine initial doripenem, imipenem, meropenem, and ertapenem dosing regimens for critically ill patients receiving prolonged intermittent RRT (PIRRT). Published body weights and pharmacokinetic parameter estimates (nonrenal clearance, free fraction, volume of distribution, extraction coefficients) with variability were used to develop a pharmacokinetic model. MCS of 5000 patients evaluated multiple regimens in 4 different PIRRT effluent/duration combinations (4 L/h × 10 hours or 5 L/h × 8 hours in hemodialysis or hemofiltration) occurring at the beginning or 14-16 hours after drug infusion. The probability of target attainment (PTA) was calculated using ≥40% free serum concentrations above 4 times the minimum inhibitory concentration (MIC) for the first 48 hours. Optimal doses were defined as the smallest daily dose achieving ≥90% PTA in all PIRRT combinations. At the MIC of 2 mg/L for Pseudomonas aeruginosa, optimal doses were doripenem 750 mg every 8 hours, imipenem 1 g every 8 hours or 750 mg every 6 hours, and meropenem 1 g every 12 hours or 1 g pre- and post-PIRRT. Ertapenem 500 mg followed by 500 mg post-PIRRT was optimal at the MIC of 1 mg/L for Streptococcus pneumoniae. Incorporating data from critically ill patients receiving RRT into MCS resulted in markedly different carbapenem dosing regimens in PIRRT from those recommended for conventional RRTs because of the unique drug clearance characteristics of PIRRT. These results warrant clinical validation.