Pharmacokinetic-pharmacodynamic model to predict drug concentrations and intraocular pressure lowering effect for a bimatoprost six-month slow-release system

Abstract

Introduction Ophthalmic drug product development and dosing regimen selection depend on animal eye drug concentration-effect relationships since human eye tissues cannot be sampled for drug quantification. This study hypothesized that a pharmacokinetic-pharmacodynamic (PK-PD) mathematical model developed based on dog studies can be applied to the human eye of different ages, based on physiological parameter adjustment, to predict drug concentrations and effects in response to a new 6-month slow-release, intracameral, intraocular pressure (IOP) lowering, anti-glaucoma bimatoprost implant. Methods Using previously reported dog concentration-effect relationship data at various doses, and the physiological parameters of dog eye, a PK-PD model was designed to predict dog aqueous humor drug concentrations and IOP lowering effects simultaneously for a given dose. After validating the model using the dog IOP data, it was applied to the human eye. Results Using a drug release rate constant of 0.0002 h−1, the model predicted the dog IOP lowering effect with an error less than 6% or less at various doses (Observed = 0.91*Predicted + 2.35; R2 = 0.98). Considering literature reported aqueous humor volumes and flow rates in old (over 60 years) and young (20 to 30 years) humans, aqueous humor elimination rate constant was estimated to be 0.9 and 0.68 h−1, respectively. The model when modified using the older human eye parameters, predicted the IOP lowering effects reported in a clinical trial with 63-year-old adults, with an error of 6.2% or less. The model, when used for young adult eye not previously tested in clinical trials, predicted lower drug concentrations and effects, possibly due to 54% higher aqueous humor volume relative to older adults. The model predicted an IOP reduction of 26.3 and 30.6%, at 10 and 15 microgram doses, respectively, in young adults. Conclusions The PK-PD model developed is useful for product design and patient dosing by predicting eye drug concentration and effect time-courses in response to implant administration at various doses, frequencies, and release rates.