V.E. Bukkems (1) , T.M. Post (2) , A. Colbers (1) , D.M. Burger (1), E.M. Svensson (1,3).
(1) Department of Pharmacy, Radboud university medical center and Radboud Institute for Health Sciences (RIHS), Nijmegen, The Netherlands. (2) Leiden Experts on Advanced Pharmacokinetics and Pharmacodynamics (LAP&P), Leiden, The Netherlands. (3) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden
Objectives: A novel formulation raltegravir was granted market authorization in 2017. This once-daily (QD) regimen of two 600 mg tablets offers an easier treatment option, has a less erratic absorption and a higher bio-availability as the former formulation containing 400mg of raltegravir. Raltegravir is a preferred agent in pregnant women, but there is no pharmacokinetic data of the 600mg formulation in pregnancy. This study aims to develop a population pharmacokinetic (popPK) model to predict the profile of the new formulation in pregnant women using simulations. The preliminary popPK model is presented here.
Methods: We combined rich pharmacokinetic data of 6 trials with Dutch healthy and HIV-infected non-pregnant subjects treated with the 400mg formulation (n=132), 1 trial with European HIV-infected pregnant women treated with the 400mg formulation (n=22) and 5 trials with American healthy and HIV-infected non-pregnant subjects treated with the 600mg formulation (n=97), to build a popPK model using NONMEM 7.4. One- to three- compartment models were tested. First- and zero-order (double) absorption models, entero-hepatic and transit absorption models were evaluated to describe the variable absorption of RAL. All flow parameters and volumes were scaled according allometric theory, with fixed allometric exponents of 0.75 for flow parameters and 1 for volumes of distribution. For pregnant women, postpartum weight was used since allometric scaling for pregnant women has not been established and confounds with the pregnancy covariate. Covariates were tested using a forward inclusion and backward elimination approach. Pregnancy was tested on apparent clearance (CL/F), apparent central volume of distribution (Vc/F), absorption rate (Ka), mean absorption time (MAT) and bio-availability (F). Values below lower limit of quantification (LLOQ) were added as LLOQ divided by 2, except consecutive values in the elimination phase. Model selection was based on the objective function value, physiological plausibility and diagnostic plots.
Results: A two compartment model with three transit absorption compartments described the data well [PK: MAT 0.302, KA 0.7, Vc/F 49, Cl/F 58.2, Q/F 6.18, Vp/F 105, RSE% for all parameters < 35%]. Interindividual variability was included on Vc/F, CL/F, Q/F and Vp/F with correlation between CL/F, Q/F and Vp/F, and inter-occasion variability was included on F and MAT. The residual error structure was proportional with a different proportional residual error for the first 3 hours after dose. The 600mg formulation had 29% (95% confidence interval; 95%CI 11-47) higher bio-availability, low-fat food decreased the bio-availability by 46% (95%CI 37-56) and food intake increased the transit time by 189% (95%CI 110-268). Pregnancy was found to be a significant covariate for F (ΔOFV -20.5) and CL (ΔOFV -17.2), while inclusion of both led to unprecise estimations. Therefore, a pregnancy covariate was included on F; pregnancy decreased F by 46% (95%CI 24-68). A posterior predictive check showed that the model was able to predict Ctrough (associated with efficacy) adequately: median (IQR) observed and simulated Ctrough in patients receiving 1200mg QD were 0.028 (0.017-0.044) mg/L and 0.031 (0.020-0.046) mg/L, respectively.
Conclusion: The popPK model adequately described the pharmacokinetics of the 400mg and 600mg raltegravir formulation in healthy, HIV-infected and pregnant subjects (400mg only). The model will be used to predict the exposure of RAL 1200mg QD in pregnant women.
Reference: PAGE () Abstr 9582 [www.page-meeting.org/?abstract=9582]
Poster: Drug/Disease Modelling - Infection