(M-092) Integrated Quantitative Systems Pharmacology and Pharmacometric Model to Evaluate Effective Buprenorphine Induction Treatment Strategies in the Era of Synthetic Opioids

Anuraag Saini, M Tech – Scientist, Vantage Research; Komalapriya Chandrasekaran, PhD – Scientist, Vantage Research; Maithreye Rengaswamy, PhD – Scientist, Vantage Research; Andreas Noack, PhD – Head of Product Engineering, Pumas AI; Jessica Wojciechowski, PhD – Director, Clinical Pharmacology & Pharmacometrics, Pumas AI; Netravat Pendsey, BS-MS – Scientist, Vantage Research; Prasad Purohit, PhD – Senior Clinical Pharmacology Scientist, Indivior Inc; Terry Horton, MD – VP, Patient Insights & Advocacy, Indivior Inc.; Mark Greenwald, PhD – Professor and Associate Chair for Research, Wayne State University

Objectives: Buprenorphine treatment induction has been associated with an increased risk of precipitated withdrawal (PW) in chronic fentanyl users diagnosed with opioid use disorder (OUD). Because PW can discourage initiation or continuation of buprenorphine treatment in patients with OUD, we developed an integrated quantitative systems pharmacology (QSP) and non-linear mixed effects model to understand the factors responsible for the increased incidence of PW in chronic fentanyl users and evaluate buprenorphine induction strategies that minimize PW.

Methods: A QSP model was developed for buprenorphine and fentanyl, integrating information on opioid pharmacokinetics (PK), mu-opioid receptor (MOR) binding kinetics, MOR tolerance mechanisms and MOR agonist activity. Prior population PK models for intravenous (IV) fentanyl and buprenorphine (IV, sublingual, and subcutaneous extended-release formulations) were incorporated to predict drug plasma exposure. The fentanyl PK model was adapted to include an adipose tissue compartment to investigate the impact of delayed release of lipophilic fentanyl on PW in chronic fentanyl users. The QSP model was calibrated using published data on respiratory depression measured after IV administration of fentanyl and buprenorphine, either alone or in combination. The relationship between opioid agonist activity (predicted by the QSP model) and withdrawal symptoms measured in patients with the Clinical Opiate Withdrawal Scale (COWS) was estimated by non-linear mixed effects modeling, using data from a clinical study of buprenorphine induction in patients with fentanyl-positive urine tests.

Results: The integrated modeling approach, leveraging mechanistic understanding from QSP and statistical flexibility of pharmacometric methodologies, successfully described changes in COWS scores in patients with and without PW following initiation of buprenorphine. The model suggests the occurrence of PW is related to an abrupt decrease in MOR agonist activity when buprenorphine is initially administered. Model diagnostics such as visual predictive checks demonstrated that the model adequately predicted the time-course in COWS scores and incidence of PW.

Conclusion: An integrated model characterizing the effects of fentanyl and buprenorphine on MORs, respiratory depression, and COWS scores was developed. This model enables the design and evaluation of improved buprenorphine induction protocols to reduce the risk of PW in chronic fentanyl users while providing adequate MOR agonist activity to motivate patient continuation in maintenance treatment.

Disclosures: Financial Support provided by Indivior Inc. The results in this abstract have been previously presented in part at CPDD, Montreal, 15-19 June 2024.

Citations: NA