Pharmacokinetic-Pharmacodynamic model for Acute Intermittent Porphyria in porphyric mice treated with a new recombinant PBGD protein
Diego Vera-Yunca (1, 2), Irantzu Serrano-Mendioroz (3), Iñaki F. Trocóniz (1, 2), Antonio Fontanellas (3), Zinnia P. Parra-Guillén (1, 2)
(1) Pharmacometrics & Systems Pharmacology; Pharmaceutical Chemistry and Technology Department; School of Pharmacy and Nutrition; University of Navarra, Pamplona, Spain. (2) IdiSNA, Navarra Institute for Health Research; Pamplona, Spain. (3) Hepatology Program, Centre for Applied Medical Research, University of Navarra, Pamplona, Spain.
Objectives: Acute intermittent porphyria (AIP) is a rare autosomal dominant disorder caused by a genetic mutation that reduces the hepatic activity of the porphobilinogen deaminase enzyme (PBGD), the third enzyme in the heme biosynthesis pathway. Precipitating factors lead to acute attacks associated to the accumulation of the neurotoxins 5-Aminolevulinic Acid (ALA) and Porphobilinogen (PBG) [1]. As the current standard-of-care drug, hemin, causes several side effects to chronical patients, new therapies are needed. In a previous work [2] we developed a data-driven disease model capable of describing the time course of excreted amounts of heme precursors in urine of porphyric mice during induced acute attacks. In this project, we aimed to refine the existing disease model to account mechanistically for known autoregulation aspects of the heme pathway, and to develop a more mechanistic pharmacokinetic-pharmacodynamic (PKPD) model using a new recombinant PBGD protein.
Methods: Acute attacks were induced at day 1, 9 and 30 in male AIP mice by intraperitoneal injection of four daily increasing doses of phenobarbital (75, 80, 85 and 90 mg/kg). The PBGD variant was administered at day 2 as a single dose intravenously at two dose levels: 60 and 300 nmol/kg. 24-h urine was collected from mice (n=27) and ALA, PBG and total porphyrins (tPOR) were quantified in control and treated animals. A total of 334 ALA, 338 PBG and 307 tPOR measurements were analyzed. As phenobarbital concentrations were not measured, a PK model was adapted from the literature to simulate phenobarbital concentrations [3]. It was assumed that phenobarbital was completely and instantly absorbed after an intraperitoneal administration. Regarding the PK of the PBGD variant, enzymatic activity data in plasma (n=16, 154 measurements) was used as a surrogate marker of the drug concentrations to build the model. Data was analyzed using the population approach with NONMEM 7.3 software, and Berkeley-Madonna was used to test different feedback mechanisms.
Results: The final disease model assumed that excreted amounts of biomarkers were dependent on the amounts of their respective biomarkers in liver and blood, represented by virtual circulating compartments with arbitrary values of 1 at baseline. Circulating levels of ALA and PBG were considered the precursors of circulating PBG and tPOR, respectively. Phenobarbital concentrations increase circulating ALA synthesis in a linear way. To acknowledge the limited activity of the endogenous PBGD enzyme to transform PBG into porphyrins, the transit between circulating PBG and circulating tPOR was modelled using a Michaelis-Menten process [Vmax equal to 1.21 (arbitrary units/h)]. In addition, a negative feedback of circulating PBG amounts on the transit between circulating ALA and circulating PBG -caused by steric hindrances on the enzyme that catalyzes the reaction from ALA to PBG- was implemented. Regarding the PKPD model, PBGD pharmacokinetics was well described using a two-compartment model with linear elimination. Drug effect was estimated using data for the low PBGD dose of 60 nmol/kg, assuming that recombinant PBGD linearly increases the endogenous enzyme maximum capacity [SLP_PBGD, the parameter governing the linear effect, equal to 9.7e-05 µL x h/pmol uroporphyrins (URO)], in agreement with its known biological action. This model, however, under-predicted the observed PBGD effect for the dose of 300 nmol/kg, indicating that an additional mechanism was needed. This issue was solved by the incorporation of an additional delayed drug effect (Emax effect, with a C50 equal to 0.20 pmol URO/µL x h) representing the PBGD activity in the liver–PBGD concentrations in liver are known to be significant only for high PBGD doses–. The final disease PKPD model was able to satisfactory describe the median values and dispersion of the data as confirmed from the goodness-of-fit and visual predictive checks.
Conclusions: A more mechanistic pharmacokinetic-pharmacodynamic model for acute intermittent porphyria in porphyric mice has been built. This model successfully describes the time course of urinary data from control mice and treated mice with the new recombinant protein for both dose values tested. This model provides a mechanistic framework to explore the impact of new therapies for acute intermittent porphyria and to extrapolate preclinical results to help taking informed decisions about dosing schemes in phase I clinical trials.
References:
[1] Karim, Z., Lyoumi, S., Nicolas, G., Deybach, J. C., Gouya, L., & Puy, H. (2015). Porphyrias: A 2015 update. Clinics and Research in Hepatology and Gastroenterology, 39(4), 412–425. https://doi.org/10.1016/j.clinre.2015.05.009
[2] Vera-Yunca, D., Serrano-Mendioroz, I., Sampedro, A., Jericó, D., Trocóniz, I. F., Fontanellas, A., & Parra-Guillén, Z. P. (2018). Computational disease model of phenobarbital-induced acute attacks in an acute intermittent porphyria mouse model. Molecular Genetics and Metabolism. https://doi.org/10.1016/J.YMGME.2018.12.009
[3] Iven, H., & Feldbusch, E. (1983). Pharmacokinetics of phenobarbital and propylhexedrine after administration of barbexaclone in the mouse. Naunyn Schmiedebergs Arch Pharmacol, 324(2), 153–159. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/6139756
