Background: TL-895 is a highly potent, orally available, selective, covalent inhibitor of Bruton tyrosine kinase (BTK) and bone marrow tyrosine kinase X-linked (BMX), under development for the treatment of myelofibrosis and chronic lymphocytic leukemia. TL-895 is given with food to increase exposure and decrease PK variability. TL-895 has lower solubility at higher pH, gastric acid reducing agents (ARAs) may decrease TL-895 exposure.
Aims: To develop and validate (i) a mechanistic absorption and physiologically-based pharmacokinetic (MA-PBPK) model to compare food and ARA effects on TL-895 PK, (ii) elucidate TL-895 absorption mechanisms, and (iii) predict PK for various dosing regimens.
Methods: A PBPK model for TL-895 was developed in GastroPlus® version 9.8.1 (Simulations Plus, Inc.) using physicochemical properties experimentally measured in vitro, or predicted from the chemical structure using the ADMET Predictor module, and refined with PK data from a crossover healthy subject fasted (reference)/fed study of 300 mg TL-895 immediate release tablets (NCT03297983) (N=18) that indicated a ~2-fold increase in AUC0-t and ~1.7 fold increase in Cmax with food.
The model was validated against a N=26, four period, healthy subject study, TL-895-204, which compared a 150 mg TL-895 tablet dose given with a low-fat meal (LFM; – reference), a high fat meal (HFM), a staggered (-10 h and +2 h) steady state (SS) 20 mg BID famotidine dose with a LFM, and SS 40 mg QD omeprazole with a LFM.
Results: Simulated TL-895 PK profiles were within 0.8-1.25 fold of observed mean fasted and fed data in both clinical studies. Model parameters indicated lower fraction absorbed (Fa, 36.3 vs. 80.3%) and absolute bioavailability (F, 9.3 vs. 26.2%) in the fasted state, relative to fed, resulted from rapid dissolution at fasted gastric pH ~1.3, followed by precipitation in intestine, where pH increased to ~6.0 and higher. At 150 mg, there was no meaningful difference in F between a HFM and LFM, or with a staggered dose of famotidine given with a LFM (F ~ 20%).
The model indicated lower Fa (55.4 vs. 85.2%) and F (7.85 vs 20.4%) in fed state with omeprazole, compared to fed state alone. Typical fed state gastric pH is ~4-5 (Parrott, 2020). As such, the gastric pH raising effect of omeprazole did not account for lower dissolution and lower exposure. To capture omeprazole effects on fed state TL-895 PK, lower intestinal bile salt concentrations (Cahan, 2006) and longer stomach transit time model parameters were necessary (Table).
Virtual population simulations for TL-895 administration with a LFM indicated relatively linear increases in Cmax and AUC for twice- or once-daily TL-895 over a dose range of 200-900 mg/day, whereas fasted dosing simulations showed Cmax and AUC peaked at 300 mg/day and declined at higher doses (Figure).
Summary/Conclusion: Physiological factors underlying higher exposure of TL-895 with fed administration were inferred from a MA-PBPK model. Simulations revealed dose linear increases in exposure with fed administration were aided by higher intestinal bile salt concentrations under fed conditions. Absorption was precipitation-limited under fasted conditions.
Model parameters indicated that lower TL-895 exposure when given with omeprazole in a fed state involves longer stomach transit time and decreased intestinal bile salt secretion. Staggered famotidine administration with food did not decrease TL-895 exposure and provides an alternative to proton pump inhibitors in patients that require an ARA.
