Intravitreal injections of vancomycin (VCM) and ceftazidime (CAZ) are commonly used to treat infectious endophthalmitis. When patient cases require retinal detachment with silicone oil (SO) tamponade, the antibiotic doses are empirically reduced to 25 %. Currently, there is no scientific evidence for these empirical dose reductions. The purpose of the present study is to determine the quantitative impact that SO tamponades have on intraocular VCM pharmacokinetics. Because of high invasiveness of frequent sampling of intraocular VCM concentrations in human, this pharmacokinetic study was performed in cynomolgus monkey's eyes. Population pharmacokinetic modeling and simulation were performed using 75 different intraocular VCM concentrations obtained from 8 male cynomolgus monkeys. A one-compartment model with a first-order diffusion rate was used as a structural pharmacokinetic model. From the covariate analysis, SO tamponade significantly decreased the volume of distribution while pars plana vitrectomy with lensectomy (PPV) significantly increased the clearance and diffusion rate constants. From the Monte Carlo simulation (n = 1,000), the median time above minimum inhibitory concentration (T>MIC, a therapeutic effect index) durations of SO and normal eyes at clinical doses of 1,000 µg were 2.6 and 11.0 days, respectively. Using intravitreal injections of VCM with SO tamponade or PPV may reduce the therapeutic effect.
Keywords: Infectious endophthalmitis; Lensectomy; Pars plana vitrectomy; Population pharmacokinetics; Silicone oil tamponade; Vancomycin.
Copyright © 2021 Elsevier B.V. All rights reserved.