Orphélie Lootens (1,2,3,4), Marthe De Boevre (1,3,4), Elke Gasthuys (2), Jan Van Bocxlaer (2), Sarah De Saeger (1,3,4,5), An Vermeulen (2)
(1) Centre of Excellence in Mycotoxicology and Public Health, Ghent University, Department of Bioanalysis, Ghent, Belgium, (2) Laboratory of Medical Biochemistry and Clinical Analysis, Ghent University, Department of Bioanalysis, Ghent, Belgium, (3) MYTOX-SOUTH®, International Thematic Network, Ghent, Belgium, (4) Cancer Research Institute Ghent (CRIG), Ghent, Belgium, (5) Department of Biotechnology and Food Technology, University of Johannesburg, Gauteng, South Africa
Objectives:
- Characterize a population
- Gather CYP450 enzyme activity phenotype data from SSA
- Subdivision of SSA into regions and statistically compare CYP450 phenotypes
- Predict and compare PK parameters across the determined regions using SimCYP
Methods:
A literature search was performed to define guidelines from EMA and FDA on PBPK populations. A systematic review was performed using EMBASE, MEDLINE and PubMed to collect CYP450 phenotype data for the SSA population. Phenotype data for various countries were compiled per CYP450 enzyme and classified into poor metabolizer (PM) or extensive metabolizer (EM) categories. Intermediate metabolizer (IM) frequencies were added to the PM group, ultrarapid metabolizers (UM) data were incorporated into the EM group. Next, 5 populations (SSA, West, East, Central and South) were built using SimCYP starting from the SouthAfrican_Population FW_Custom population and modified with the CYP450 frequency data per region. [1] A pairwise proportions test with Bonferroni correction was performed using R studio on the phenotype frequency data across the 5 regions per CYP450 enzyme. In case of significance, a cohen’s d test to verify for standardized mean differences was applied. Next, simulations in the 5 populations (SSA, West, East, Central and South) with 5 probe substrates were performed in SimCYP using the standard care dosing, i.e. 600 mg efavirenz (CYP2B6), 10 mg warfarin (CYP2C9), 200 mg mephenytoin (CYP2C19), 22 mg dextromethorphan twice daily (CYP2D6) and 5 mg midazolam (CYP3A5). The population size was 1,000 healthy volunteers (20-50 y/o; proportion of males: 50 %) and the treatment period was 30 days. Bar charts were plotted for each region for following PK parameters i.e. Cmax, Tmax, AUC, CL.
Results:
Guidelines of the EMA and FDA provided following information. In the qualification of populations within PBPK models, the process involves the establishment of system-dependent parameters, followed by the prospective prediction of PK parameters within the target population for verification. Phenotype data was found on CYP2B6, CYP2C9, CYP2C19, CYP2D6 and CYP3A5. The statistical testing revealed significant differences for CYP2B6 between SSA and West (p = 7.2e-11), SSA and East (p < 2e-16), West and East (p = 0.0061), West and South (p = 6.9e-11) and South and East (p < 2e-16). For CYP2C19 there was a statistical difference between West and East (p = 0.0075). For CYP2D6 a statistical difference was observed for SSA and Central (p = 5.7e-5), East and Central (p = 9.1e-5) and for South and Central (p = 2.8e-6). The cohen’s d values were between 0.4 and 0.5, indicating a moderate difference. For CYP2C9 and CYP3A5, no significant differences were observed. The statistically significant differences could also be visually observed from the SimCYP prediction bar charts. Furthermore, since the population models were built starting from an available population, more differences were observed in the bar charts. The statistical testing was only performed on the data gathered from literature whereas the PBPK simulations also considered other CYP450 frequencies from the available South African population.
Conclusions:
PBPK population development is not regulated, only guided by the EMA and FDA. Phenotype data on CYP450 enzymes can be collected for subregions within an already existing population. Statistical analysis can be performed to check for significant differences. Numerous studies have substantiated the presence of inter-ethnic diversity within SSA [4], [5]. This study justifies the subdivision of SSA into West, East, Central and South Africa. Ideally, subpopulations are verified using available in vivo data. Further research should be performed to investigate if further subdivisions in these regions should be performed.
References:
[1] D. Namanya, Judith Wanyama, N. Haghtalab, I. N. Djenontin, T. Bilintoh, and L. Mungai, “African Unit 12 – Geography of Africa.” https://aac.matrix.msu.edu/modules/african-modules/african-unit-12-geography-of-africa/
[2] EMA, “Guideline on the reporting of physiologically based pharmacokinetic (PBPK) modelling and simulation,” vol. EMA/CHMP/4, no. December, pp. 3–15, 2018, [Online]. Available: www.ema.europa.eu/contact%0Ahttps://www.ema.europa.eu/en/documents/scientific-guideline/adopted-reflection-paper-use-extrapolation-development-medicines-paediatrics-revision-1_en.pdf
[3] U. S. F. & D. Administration, “Physiologically Based Pharmacokinetic Analyses — Format and Content,” Guidance, no. August, 2018.
[4] S. C. Schuster et al., “Complete Khoisan and Bantu genomes from southern Africa,” Nature, vol. 463, no. 7283, p. 943, Feb. 2010, doi: 10.1038/NATURE08795.
[5] S. A. Tishkoff et al., “The Genetic Structure and History of Africans and African Americans,” Science, vol. 324, no. 5930, p. 1035, May 2009, doi: 10.1126/SCIENCE.1172257.
Reference: PAGE 32 (2024) Abstr 10895 [www.page-meeting.org/?abstract=10895]
Poster: Drug/Disease Modelling - Absorption & PBPK