In Vitro Inhibitory Effects andMechanisms of Tyrosine Kinase Inhibitors on Warfarin Metabolism

Introduction: Tyrosine kinase inhibitors (TKIs) and the anticoagulant
drug warfarin are commonly co-prescribed in patients with cancerassociated
venous thromboembolism. Many TKIs have been suspected of causing hazardous drug-drug interactions (DDIs) with warfarin on the
basis of case studies, but the biochemical mechanisms causing these
interactions have not been demonstrated.
Objectives: As inhibition of cytochrome P450 (CYP) enzymes is one of
the central DDI mechanisms, we investigated 11 TKIs for their inhibitory
effects on the major enzymes involved in the metabolism of warfarin
enantiomers.
Methods: An in vitro cocktail assay with three CYP enzymes was applied
in this study. Diclofenac 4'-hydroxylation, midazolam 1'-
hydroxylation and tacrine 1-hydroxylation were used as marker reactions
for CYP2C9, CYP3A and CYP1A2 activity, respectively.
Results: After a 30-min preincubation with NADPH to screen for possible
time-dependent inhibition, trametinib, alvolcidib and cediranib exhibited
an increased inhibition of CYP3A, tozasertib an increased inhibition
of both CYP3A and CYP2C9, and linsitinib and masitinib an increased
inhibition of both CYP3A and CYP1A2, compared to experiments with
no preincubation. In detailed experiments, tozasertib and linsitinib were
identified as time-dependent inhibitors of CYP3A, with inhibitor concentration
that supports half-maximal rate of inactivation (KI) and maximal
inactivation rate (kinact) of 400 μM and 0.026 min-1, and 231μM and
0.056 min-1, respectively. Linsitinib was the only inhibitor causing a
time-dependent inhibition of CYP1A2 with KI and kinact of 9.06 μM
and 0.017 min-1. However, trametinib, alvolcidib and cediranib were
negative for significant time-dependent inhibition of CYP3A.
Moreover, the competitive inhibitory effects of linsitinib, tozasertib and
trametinib towards CYP2C9were strong,with a Ki of 1.43, 2.85 and 0.56
μM, respectively. Masitinib and vatalanib exhibited competitive inhibition
of both CYP3A (Ki=1.33 μM, Ki=0.26 μM) and CYP2C9 (Ki=1.98
μM, Ki=0.30 μM). Vatalanib also noncompetitively inhibited CYP1A2
with a Ki of 1.96 μM.
Conclusion: The results suggest that several of the TKIs studied may
cause DDIs by inhibition of CYP1A2, CYP2C9 or CYP3A4. More attention
should be paid on to avoid unnecessary clinical DDI risks when
such TKIs are co-administrated with warfarin.