(M-029) Bilirubin Elevation Due To Atazanavir Treatment May Primarily Be Mediated By Inhibition Of OATP1B1/3 Rather Than That Of UGT1A1: PBPK Analysis Of Bilirubin-Atazanavir-Polymorphism Interaction

Jin Dong, Ph.D. – Associate Director, AstraZeneca; Pradeep Sharma, Ph.D. – Director, AstraZeneca; Rasha Emara, B.S. – Student, University of the Pacific; Derek Cheung, Pharm.D. – Student, University of the Pacific; Weifeng Tang, M.D., Ph.D. – Executive Director, AstraZeneca; Diansong Zhou, Ph.D. – Senior Director, AstraZeneca; David Boulton, Ph.D. – Executive Director, AstraZeneca; Mats Någård, Ph.D. – Executive Director, AstraZeneca; Miki Park, Ph.D. – Professor, University of the Pacific

Objectives: atazanavir has been reported to increase total bilirubin (TB) level up to 9-fold [1]. It is widely believed that the observed TB elevation is primarily due to UGT1A1 inhibition. However, OATP1B1/3 may contribute to the interaction as atazanavir is a potent OATP1B1/3 inhibitor and the hepatic uptake of both UB and conjugated bilirubin (CB) are mediated by them [2]. The objective of this study was to develop physiologically based pharmacokinetic (PBPK) models to quantitatively understand the contribution of UGT1A1 and OATP1B1/3 to the bilirubin and atazanavir interaction.

Methods: All PBPK modeling was conducted in Simcyp (version 21). The UB and CB models were modified from a published model [3]. The OATP1B1/3 and UGT1A1 kinetics were validated with clinical interaction with rifampicin and the observed baseline levels in subjects with UGT1A1 variants, respectively. The atazanavir and ritonavir models were modified from the built-in Simcyp models. Their effects on UGT1A1 and OATP1B1/3 were validated with clinical data of drug-drug interaction (DDI) with raltegravir, atorvastatin or pravastatin. The darunavir model was constructed based on publications [4, 5]. The cobicistat model was developed and validated with its clinical pharmacokinetics (PK) and DDI with midazolam. Then, the models were validated with observed TB, UB, and/or CB elevation in the atazanavir boosted or unboosted treatments in healthy volunteers or subjects with UGT1A1 variants [6]. Finally, the TB and UB elevation was simulated assuming without either UGT1A1 or OATP1B1/3 inhibition to assess the contribution of UGT1A1 and OATP1B1/3 to the interaction.

Results: Most of the simulated PK, DDI, and bilirubin interactions were within 0.8-1.25-fold and all were within 2-fold. The model reasonably simulated UB increase (1.87, 2.90, 3.73 mg/dL) with atazanavir/ritonavir treatment compared with the observed values of 1.79, 3.04, 5.00 mg/dL in extensive (EM), intermediate (IM), and poor metabolizers (PM) of UGT1A1 [5]. The simulated 2.2-fold elevation of CB and 39% reduction of percentage of CB in TB were consistent with the observed values of 2.0-fold and 56% [7], respectively.
The simulated percent change of TB was 471% with atazanavir/ritonavir in EM whereas it reduced to 345% and 37% without UGT1A1 or OATP1B1/3 inhibition, respectively. Similarly, it was shown that OATP1B1/3 inhibition has higher impact on the interaction in IM and PM than UGT1A1 inhibition.
When comparing the sensitivity of UB and raltegravir as UGT1A1 substrates, the simulated AUC ratios of UB and raltegravir due to the UGT1A1 inhibition by atazanavir were 1.26 and 1.78, respectively.

Conclusions: The PBPK modelling suggested that bilirubin and atazanavir interaction may primarily be driven by OATP1B1/3 inhibition. UB is less sensitive than raltegravir upon UGT1A1 inhibition.

Citations: [1] Dekker, D., et al. Arterioscler. Thromb. Vasc. Biol., 2011, 458
[2] Miners, J.O., et al. Pharmacol. Ther., 2023, 108459
[3] Dong, J., et al. PBPK162. ACoP12
[4] Dong, Z., et al. CPT Pharmacometrics Syst. Pharmacol., 2020, 456
[5] Wagner, C., et al. J. Clin. Pharmacol., 2017, 1295
[6] Anderson, P.L., et al. J. Antimicrob. Chemother., 2009, 1071
[7] Roy‐Chowdhury, J., et al. Clin. Pharmacol. Drug Dev., 2017, 140