Aziz Chaouch (1), Youssef Daali (2), Edoardo Baglivo (3), Laura Cottet (3), Jules Desmeules (2), Marianne Gex-Fabry (4), Chantal Csajka (1)(5)
(1) Division of Clinical Pharmacology, University Hospital and University of Lausanne, Lausanne, Switzerland, (2) Service of Clinical Pharmacology and Clinical Toxicology, University Hospital of Geneva, Geneva, Switzerland (3) Department of Ophthalmology, University Hospital of Geneva, Geneva, Switzerland (4) Department of Psychiatry, Geneva University Hospitals, Geneva, Switzerland (5) School of Pharmaceutical Sciences, University of Geneva, Geneva, Switzerland
Objectives: To model drug concentration profiles in plasma and in the aqueous humour of patients who underwent cataract surgery, following a single oral administration of 200mg voriconazole (VRC), and to assess the effect of demographic (body weight, age, gender) and phenotypic covariates (CYP2C9, CYP2C19 and CYP3A activities) on drug exposure.
Methods: Concentrations in plasma (2-3 samples per subject) and in the aqueous humour (1 sample per subject) were obtained from 33 patients (16 males, 17 females) aged 48-100y over a period of 24 hours after administration of a single oral dose of 200mg VRC. Concentration profiles were analyzed on the log-scale using a non-linear mixed-effects model (NONMEM v.7.3). The volume of distribution of the aqueous humour was fixed to 200uL according to the true physiological intracameral volume [1]. Potential influential covariates were screened using plots of Bayesian estimates of random effects versus covariates whenever shrinkage was acceptable (<30%). A stepwise forward selection procedure using a level alpha=0.05 (delta OF>3.84) was then used to incorporate covariates in the model, followed by a backward elimination procedure using a level alpha=0.01 (delta OF>6.63). Goodness of fit was assessed using standard graphical diagnostic plots as well as visual predictive checks.
Results: The data were adequately described by a 2 compartment model describing VRC in plasma and an additional compartment for characterizing concentrations in the aqueous humour with linear absorption and linear elimination. The high inter-individual variability of clearance (CV 82.1%) and volume of distribution (CV 67.9%) of the central compartment could be partly explained by covariates: clearance and central volume of distribution were influenced by body weight and fast metabolizers of CYP2C19 had higher clearances after adjusting for body weight. The coefficient of penetration of voriconazole into the aqueous humour was estimated at 50.7%.
Conclusions: VRC is characterized by an important intersubject variability that could be partially explained by body weight and CYP2C19 activity. Orally administered VRC seems to achieve therapeutic aqueous levels in the non inflamed human eye. Simulations of intraocular penetration while considering the high variability in VRC pharmacokinetics will be performed to evaluate the proportion of patients with levels above the inhibitory concentrations of most frequently encountered mycotic species involved in fungal endophtalmitis.
References:
[1] Toris CB, Yablonski ME, Wang YL et al. Aqueous humor dynamics in the aging human eye. Am J Ophtalmol (1999), 127:407-412
Reference: PAGE 23 () Abstr 3146 [www.page-meeting.org/?abstract=3146]
Poster: Drug/Disease modeling - Infection