(M1130-03-18) Model-Informed Approach for Formulation Development of Extended-Release Phenserine Tablets to Mitigate pH-Dependent Degradation

Purpose: Phenserine is a centrally-active, non-competitive, selective inhibitor of acetylcholinesterase (AChE). Clinical trials have shown phenserine to be safe and well tolerated at a single oral dose of 10 mg, however, dose-limiting side effects were observed with a 20 mg dose. Phenserine is primarily metabolized by CYP3A4, has a short plasma half-life, and shows non-linear pharmacokinetics with increased drug exposure at higher doses. In a Phase III trial, phenserine did not meet established efficacy targets, but the clinical failure of the drug was later attributed to inadequate drug exposure throughout the dosing interval. Improved phenserine efficacy may be achieved using an extended-release dosage form, and a 12 h extended-release formulation was developed containing 20 mg of phenserine in an HPMC matrix tablet. During formulation development, phenserine was observed to form a colored degradant in near-neutral and alkaline pH environments, and at pH 7, which is representative of the pH in the distal small intestine, the half-life was estimated to be only 5 h (40 °C). To limit chemical degradation in the intestinal lumen resulting in reduced bioavailability, a gastroretentive tablet composed of a polyethylene oxide-xanthan gum matrix was developed. Optimization of the drug release profile from the gastroretentive tablet to maximize bioavailability and minimize luminal degradation could benefit from additional experimental and in silico evaluations, and a PBPK model was developed to assist in the identification of effective doses and delivery rates from gastroretentive tablets able to achieve once or twice daily dosing.

Methods: GastroPlus® (version 9.8.3, SimulationsPlus Inc, USA) and its PBPKPlus® module were used to develop a whole-body PBPK model for phenserine. The Population Estimates of Age Related (PEAR) physiology module was used to match the physiologies described in the published clinical reports. Physiochemical and other parameters were obtained from literature, in silico predictions (ADMET Predictor version 10.4), and from drugs similar to phenserine in structure and function. The brain:plasma partition coefficient (Kp_brain) reported from preclinical investigations was included in the model and was also used to describe the partitioning in other fatty tissues including adipose and yellow marrow. The intrinsic clearance parameters, V_max and K_m, for CYP3A4 activity in the gut and liver were optimized to achieve saturation at high oral doses (20 mg). Chemical degradation of phenserine was also incorporated into the distal GI lumen compartments and in the blood. The PBPK model was used to predict the plasma concentration profiles for the previously developed extended-release and gastroretentive tablets. Further simulations were used to inform the design of additional gastroretentive tablets, and various release profiles and doses were investigated using a gastric retention of four hours under fed state conditions.

Results: The PBPK model shows a good predictive capability to describe the reported clinical results, especially over the initial 12 hours post-administration. The model showed saturation of metabolism at a higher dose (20 mg), and the PK parameters predicted by the model were within a two-fold range of the parameters obtained from a non-compartmental analysis of the clinical data. Simulations based on the drug release from the 20 mg extended-release tablet or the prototype gastroretentive tablet did not predict the maintenance of therapeutic plasma concentrations for a 12 hour interval. Simulations using a 60 mg phenserine dose in a gastroretentive tablet were able to sustain therapeutic drug concentrations for an extended period of time while maintaining concentrations below those previously associated with significant side effects.

Conclusion: A PBPK model for phenserine capturing the chemical degradation of the drug in the GI lumen and blood along with saturation of CYP3A4 intestinal and hepatic metabolism was developed. The model was used to identify drug doses and release profiles able to provide for twice daily phenserine dosing to improve patient compliance while providing therapeutic concentrations and limiting the occurrence of side effects associated with the high plasma concentrations observed using immediate-release tablets.