Onset and disappearance of induction on CYP3A4 and UGT1A1 substrates using PBPK modelling
Sara Bettonte (1,2), Mattia Berton (1,2), Felix Stader (3), Manuel Battegay (1,2), Catia Marzolini (1,2,4)
(1) University Hospital Basel, Switzerland, (2) University of Basel, Switzerland, (3) Certara UK Limited, United Kingdom, (4) University of Liverpool, United Kingdom
Introduction: Contemporary antiretrovirals (ARVs) have a lower propensity to cause drug-drug interactions (DDIs), however they can be subject to DDIs notably with inducers. The coadministration with an inducer is unavoidable in some clinical situations (e.g., HIV and tuberculosis co-infection). Induction takes several days to fully develop and disappear upon initiation and discontinuation of an inducer. However, it is unknown whether the time to reach maximal induction and resolution of induction varies depending on the inducer strength, elimination half-life or degradation rate of the metabolic enzyme. This question is highly relevant given that dolutegravir, the first-line ARV recommended by WHO, is metabolized by the cytochrome P450 (CYP) 3A4 and uridine 5'-diphospho-glucuronosyltransferase (UGT) 1A1 [1] and therefore requires dose adjustment in presence of an inducer.
Objectives: This study aimed to use physiologically based pharmacokinetic (PBPK) modelling to simulate the time to reach the maximum induction effect of CYP3A4 and UGT1A1 by rifampicin, efavirenz, and rifabutin as well as the time for the enzyme abundance to return to baseline levels after stopping the administration of the inducers.
Methods: Our in-house PBPK model built in Matlab®2020a [2] was used to develop the drug models for dolutegravir (administered dose 50 mg), rifampicin (600 mg), efavirenz (600 mg), and rifabutin (300 mg). In order to reproduce steady state induction conditions and to verify the ability of the model to correctly predict DDIs, the simulations between dolutegravir and the three inducers were compared against observed data. In addition, available clinical data from treatment switch studies were used to simulate the effect of residual induction by efavirenz on the pharmacokinetics of dolutegravir or rilpivirine (metabolized by CYP3A4). A cohort of 100 virtual individuals aged 20 to 50 was generated to simulate the unknown switch scenarios (i.e., stopping rifampicin or rifabutin while initiating dolutegravir). Finally, the percentage of change in enzyme abundance levels were calculated for all three inducers during and after stopping their administration.
Results: The drug models were successfully verified as predictions were within 2-fold of observed clinical data. For the switch scenarios (i.e., from efavirenz to dolutegravir and from efavirenz to rilpivirine), the residual inducing effect of efavirenz on the pharmacokinetics of the two victim drugs was correctly predicted with simulation results falling within 2-fold of clinical observed data. For efavirenz, 14 days were necessary for the CYP3A4 enzymes to return to baseline level after stopping its administration, while the recovery of UGT1A1 was faster. For the unknown scenarios, rifampicin was predicted to decrease dolutegravir trough concentration (Cτ) by 92% and by 60% at day 1 and 7 after stopping its administration, the induction effect of rifampicin was still visible after 14 days. On the other hand, rifabutin was predicted to decrease Cτ by 26% and by 13% at day 1 and 7 after stopping its administration and only 7 days were necessary to the CYP3A4 abundance to return to baseline. To reach the maximal CYP3A4 induction, rifampicin was predicted to need at least 14 days of continuous administration, while 10 days were predicted to be sufficient for rifabutin. For both rifampicin and rifabutin, the recovery of UGT1A1 was faster than CYP3A4.
Conclusions: For the strong inducer rifampicin, 14 days were predicted to be necessary to reach maximal induction onset and to return to baseline abundance. These results are in line with previous PBPK model data [3, 4] and clinical data. The time to reach maximal induction and de-induction was predicted to vary between 7 and 14 days for moderate inducers depending on their pharmacokinetic properties. For UGT1A1, the onset and de-induction processes were predicted to be faster than CYP3A4; however, this finding needs to be interpreted with caution given that the degradation rate constant (Kdeg) was derived from a rat model as it has not been measured in humans [5]. Altogether, our findings suggest that the main drivers of induction onset and de-induction are: Kdeg, inducer half-life, inducer plasma concentration, and induction strength. Furthermore, our simulations support the common practice of maintaining the adjusted dosage of ARV for another 2 weeks after stopping an inducer.
References:
[1] U.S. Food and Drug Administration. Tivicay label 2013. Access: March 2023. Available from: https://www.accessdata.fda.gov/drugsatfda_docs/label/2013/204790lbl.pdf.
[2] Stader F et al. CPT: PSP (2019) 8, 444-459.
[3] Kinvig H et al. Conference on Retroviruses and Opportunistic Infections (2020) Abstr: 450 https://www.croiconference.org/abstract/high-dose-rifampicin-for-the-treatment-of-leprosy-in-hiv-patients-taking-dolutegravir/.
[4] Kapetas AJ et al. AAPS J (2019) 21, 78.
[5] Suzuki M et al. J Hum Genet (2014) 59, 158-62.