Physiologically based pharmacokinetic modelling to predict the concentration-time profile of an antibody-drug conjugate and hepatic and renal impairment.
Felix Stader, Cong Liu, Adriana Zyla, Abdallah Derbalah, Armin Sepp
(1) Certara UK Ltd., Certara Predictive Technologies Division, Level-2-Acero, 1 Concourse Way, Sheffield, S1 2BJ, UK
Introduction:
Antibody-drug conjugates (ADC) present a promising treatment option in oncology. Antibodies have high specificity for their cognate antigens and act as carriers for the covalently attached highly potent cytotoxic drugs. The ADC brentuximab vedotin was approved about a decade ago against the cancer of lymphatic cells. The antibody part of the ADC targets CD30 and is linked to the microtubule-disrupting agent monomethyl auristatin E (MMAE).
The prevalence of hepatic and renal impairment (HI and RI) increased among cancer patients [1]. Earlier research demonstrated that 55% and 15% of cancer patients had mild and moderate RI [2]. HI and RI are not thought to have a significant impact on the pharmacokinetics (PK) of an antibody, but they can lead to a higher exposure of the cytotoxic payload. A dose adjustment might be required for both comorbidities to ensure a safe therapy for the patient, but clinical data are sparse, which can be overcome by physiologically based pharmacokinetic (PBPK) modelling, an predictive approach for scenarios which are difficult or unethical to study in a clinical trial. PBPK modelling was successfully used to predict the PK in HI and RI for small molecule drugs [3,4]. Additionally, drug-drug interactions between the cytotoxic payload of an ADC and CYP3A inhibitors were successfully predicted [5,6].
Objectives: The aim of this proof-of-concept study was to utilize PBPK modelling to predict the impact of HI and RI on the PK of the ADC brentuximab vedotin.
Methods: A whole-body PBPK model for conjugated therapeutic proteins [7,8], implemented in the Simcyp Simulator V23® was used [9]. The example drug in this proof-of concept study was brentuximab vedotin, because scarce data (7 individuals with HI and 10 with RI) are available in the literature to evaluate the performance of the PBPK model [10]. The demographics of the study were matched as closely as possible, when setting up a clinical study with 100 participants (10 individuals in 10 trials) per patient group. The HI and RI populations, implemented in the Simcyp Simulator® were used without any changes. Brentuximab vedotin was injected as 2h intravenous infusion at a dose of 1.2 mg/kg. The predictions of the plasma concentration-time profile, the peak concentration (Cmax), and the area under the curve extrapolated to infinity (AUCinf) were compared against the clinically observed data of Zhao et al. [10]. The ratio of the impaired population to healthy volunteers and hepatic and renal impaired population for Cmax and AUCinf were calculated for the released payload to get an understanding for the impact of HI and RI on brentuximab vedotin.
Results:
The clinically observed concentration-time profile of brentuximab vedotin in unimpaired individuals was predicted within the 95% confidence interval and the investigated PK parameters were predicted within 1.25-fold of the observed data for the antibody and the released payload.
All Cmax and 85% of AUCinf values for the total antibody were predicted within 1.5-fold range of the observed data in patients with HI and RI. This demonstrated that mAb PK was not affected by the disease state.
The concentration-time profiles, Cmax and AUCinf for the released MMAE in moderate and severe HI were predicted within 1.5-fold of the observed data, demonstrating the ability of the PBPK approach to predict MMAE disposition in HI. The investigated PK parameters were under-predicted by more than 2-fold in mild HI. However, the single patient had higher concentrations of the drug in circulation compared with moderate and severe HI.
The impact of RI was over-predicted for MMAE, but the observed concentration-time data were within the lower boundaries of the simulation. Cmax and AUCinf were predicted within 2-fold of the observed data for mild, moderate, and severe RI.
The predicted exposure ratio between healthy volunteers and mild, moderate, and severe HI for the released MMAE were 1.5-, 1.8-, and 2.1-fold. MMAE exposure was predicted to increase by 1.7-, 2.1-, and 2.6-fold in mild, moderate, and severe RI.
Conclusions: This proof-of-concept study demonstrated that PBPK modelling is a useful approach to predict the impact of HI and RI on the PK of ADCs. The model estimated more than 2-fold change in AUCinf for released MMAE in severe HI as well as in moderate and severe RI, which warrants careful monitoring in clinical practice since 15% of cancer patients have a GFR below 60 mL/min.
References:
[1] Krens, S.D.; Lassche, G.; Jansman, F.G.; Desar, I.M.; Lankheet, N.A.; Burger, D.M.; van Herpen, C.M.; van Erp, N.P. Dose recommendations for anticancer drugs in patients with renal or hepatic impairment. The Lancet Oncology 2019, 20, e200-e207.
[2] Launay‐Vacher, V.; Oudard, S.; Janus, N.; Gligorov, J.; Pourrat, X.; Rixe, O.; Morere, J.F.; Beuzeboc, P.; Deray, G. Prevalence of renal insufficiency in cancer patients and implications for anticancer drug management: the renal insufficiency and anticancer medications (IRMA) study. Cancer 2007, 110, 1376-1384.
[3] Alsmadi, M.t.M.; AL‐Daoud, N.M.; Jaradat, M.M.; Alzughoul, S.B.; Abu Kwiak, A.D.; Abu Laila, S.S.; Abu Shameh, A.J.; Alhazabreh, M.K.; Jaber, S.a.A.; Abu Kassab, H.T. Physiologically‐based pharmacokinetic model for alectinib, ruxolitinib, and panobinostat in the presence of cancer, renal impairment, and hepatic impairment. Biopharmaceutics & Drug Disposition 2021, 42, 263-284.
[4] Heimbach, T.; Chen, Y.; Chen, J.; Dixit, V.; Parrott, N.; Peters, S.A.; Poggesi, I.; Sharma, P.; Snoeys, J.; Shebley, M. Physiologically‐based pharmacokinetic modeling in renal and hepatic impairment populations: a pharmaceutical industry perspective. Clinical Pharmacology & Therapeutics 2021, 110, 297-310.
[5] Samineni, D.; Ding, H.; Ma, F.; Shi, R.; Lu, D.; Miles, D.; Mao, J.; Li, C.; Jin, J.; Wright, M. Physiologically based pharmacokinetic model‐informed drug development for polatuzumab vedotin: label for drug‐drug interactions without dedicated clinical trials. The Journal of Clinical Pharmacology 2020, 60, S120-S131.
[6] Choules, M.P.; Zuo, P.; Otsuka, Y.; Garg, A.; Tang, M.; Bonate, P. Physiologically based pharmacokinetic model to predict drug–drug interactions with the antibody–drug conjugate enfortumab vedotin. Journal of Pharmacokinetics and Pharmacodynamics 2023, [Online ahead of print], 1-12.
[7] Li, L.; Gardner, I.; Dostalek, M.; Jamei, M. Simulation of monoclonal antibody pharmacokinetics in humans using a minimal physiologically based model. The AAPS Journal 2014, 16, 1097-1109.
[8] Gill, K.L.; Gardner, I.; Li, L.; Jamei, M. A bottom-up whole-body physiologically based pharmacokinetic model to mechanistically predict tissue distribution and the rate of subcutaneous absorption of therapeutic proteins. The AAPS Journal 2016, 18, 156-170.
[9] Li, L.; Chen, S.-C.; Stader, F.; Rose, R.; Rao, I.; Gardner, I.; Jamei, M.; Shen, B.-Q.; Yip, V.; Chen, Y. A whole body physiologically based pharmacokinetic model for antibody drug conjugates–model development and validation in rat. In Proceedings of the Population Approach Group Europe (PAGE), Budapest, Hungary, 2017.
[10] Zhao, B.; Chen, R.; O'Connor, O.A.; Gopal, A.K.; Ramchandren, R.; Goy, A.; Matous, J.V.; Fasanmade, A.A.; Manley, T.J.; Han, T.H. Brentuximab vedotin, an antibody–drug conjugate, in patients with CD30‐positive haematologic malignancies and hepatic or renal impairment. British Journal of Clinical Pharmacology 2016, 82, 696-705.
