Gaps in predicting clinical doses for cannabinoids therapy: Overview of issues for pharmacokinetics and pharmacodynamics modelling.

Model-based prediction on clinical doses for cannabinoids therapy is beneficial in the clinical setting, especially for seriously ill patients with both altered pharmacokinetics and pharmacodynamic responses. The objective of this article is to review the currently available PK and/or PD models of Δ9-tetrahydrocannabinol (THC) and cannabidiol (CBD) and to highlight the major issues for modelling this complex therapeutic area. A systematic search was conducted in the electronic databases PubMed and EMBASE using the key words 'cannabis', 'cannabinoid', 'tetrahydrocannabinol', 'THC', 'cannabidiol', 'CBD', 'pharmacokinetic model', 'pharmacodynamics model' and their combinations. Twelve empirical PK and/or PD models for THC for humans were identified. Among them, ten were developed from data of healthy participants and two were from ill patients. Models for CBD were not found. Model-based prediction on appropriate doses for cannabinoids therapy for ill patients is currently limited due to insufficiency of relevant PK and PD data. High-quality PK and PD data of cannabinoids for patients with different illnesses is needed for model development. Mechanism-based PK and PD models are promising for improved predictive dosing performance for ill and comorbid patients.