Stijn W. van Beek (1), P. Kellie Turner (3), Jan-Stefan van der Walt (2)
(1) Radboudumc, NL, (2) Eli Lilly and Company, UK, (3) Eli Lilly and Company, USA
Objectives:
Abemaciclib, a potent inhibitor of CDK4&6 for treatment of HR+ HER2- locally advanced or metastatic breast cancer [1,2,3], is primarily metabolised by CYP3A4 to several active metabolites (M2, M18 and M20) [4,5]. The best measure of the active moiety is the potency-adjusted unbound AUC of abemaciclib plus its metabolites. Hepatic impairment (HI) can reduce abemaciclib clearance and affect protein binding. The effects of HI on the active moiety were(1) assessed using a semi-mechanistic population PK model incorporating changes in hepatic blood flow, protein binding and intrinsic clearances, and (2) validated by comparing the Child-Pugh Score (CPS) [6] and NCI criteria [7] to categorize the severity of HI for covariate modeling.
Methods: Body weight, liver volume estimates and PK data (parent and metabolites) after a single dose of 200mg abemaciclib in healthy subjects (HS, n=10), and subjects with mild (n=9), moderate (n=10) and severe (n=6) HI based on CPS (NCT02387814) were analysed. A mechanistic population PK model developed from 12 clinical trials in cancer patients and healthy subjects (8) was used to estimate changes in drug extraction and CYP3A4-mediated metabolism (NONMEM v7.3). For covariate modeling using CPS the relevant parameter-covariate relationships were determined using HI literature and scientific plausibility. Covariate effects were estimated where possible. Alternatively, a range of covariate effect values, selected based on literature, was tested. The best models were selected based on improvement in predictive performance using visual predictive checks (PsN v4.7). For the NCI model, only the covariate relationships included in the final CPS model were tested. A sensitivity analysis was performed to select the final covariate effect values for the NCI model. The abemaciclib, metabolites and active moiety AUC for HI groups were compared to HS by ANOVA. To evaluate potential dose adjustments, steady-state concentrations after once daily (QD) or twice daily (Q12H) doses of 50, 100, 150 and 200mg in subjects with mild, moderate or severe HI were simulated and compared to the clinical efficacy target (200 ng/mL) [9].
Results:
Covariate modeling with CPS criteria resulted in inclusion of three ordinal parameter-covariate relationships: hepatic blood flow decreased 30% (mild), 44% (moderate), and 76% (severe); metabolism of abemaciclib to M2 decreased 15% (mild), 23% (moderate) and 78% (severe); and the fraction of M2 metabolized to M18 decreased 30% (mild), 50% (moderate) and 70% (severe).
These changes resulted in increased abemaciclib and decreased metabolite exposure: the active moiety increased 1.8 (95%CI 1.24-2.42) fold in severe HI, but was not different for mild and moderate HI from HS. At steady-state these changes would result in a significant increase in active moiety exposure in severe HI (2.55 fold, 95%CI 1.25-3.85) but no difference was found for mild and moderate HI. This increase was mainly driven by a 4.47-fold increase in abemaciclib exposure.
When using the NCI criteria, there were 17 HS, 8 mild HI, 7 moderate HI, and only 3 subjects with severe HI. Some individuals changed from moderate CPS category to HS status with NCI. For covariate analyses the HI severity was reclassified as either HS/mild HI or moderate/severe HI. The hepatic blood flow decreased 80%, the metabolism of abemaciclib to M2 decreased by 75%, and the fraction of M2 metabolized to M18 decreased by 78% in moderate/severe HI compared to HS/mild HI subjects, which was similar the estimates from the CPS model.
Steady-state predictions with the CPS model supported a dose adjustment to 200mg QD for the severe HI group: active moiety exposure was slightly higher (1.28-fold increase) compared to HS. Dose adjustment to 150mg QD resulted in total exposure comparable to HS (0.96 fold). Dose reductions as low as 100mg Q12H maintained trough concentrations above 200 ng/mL in severe HI.
Conclusions: Changes in liver blood-flow and reduced metabolism in severe HI are consistent with increased total active moiety of abemaciclib. Categorizing HI severity using the CPS and NCI criteria provided similar estimates of the effect of severe HI on abemaciclib PK. At steady state, the recommended dose adjustment for severe HI to 200mg QD would result in exposure of the active moiety comparable to HS and maintain the abemaciclib concentrations above the clinical threshold.
References:
[1] Sledge G.W. Journal of Clinical Oncology, 2017. 35(25): p. 2875-2884
[2] Goetz, M.P. Journal of Clinical Oncology. 0(0): p. JCO.2017.75.6155
[3] Dickler, M.N. Clin Cancer Res, 2017. 23(17): p. 5218-5224
[4] Kulanthaivel, P. Cancer Research, 2016. 76(14 Supplement): p. CT153-CT153
[5] Burke, T. Cancer Research, 2016. 76(14 Supplement): p. 2830-2830
[6] Pugh, R.N.H., et al. BJS, 1973. 60(8): p. 646-649
[7] National Cancer Institute Organ Dysfunction Working Group (NCI-ODWG), Protocol Template for Organ Dysfunction Studies, Cancer Therapy Evaluation Program (CTEP), Editor. 2018.
[8] Chigutsa E. ACoP. 2017: Fort Lauderdale, FL
[9] Tate, S.C., et al. Clin Cancer Res, 2014. 20(14): p. 3763-74.
Reference: PAGE 28 (2019) Abstr 8848 [www.page-meeting.org/?abstract=8848]
Poster: Drug/Disease Modelling - Oncology