Historically, drugs have been selected using various methods (eg, biological and chemical screens). Candidate drugs were often pushed into the clinic with only a rudimentary understanding of the link between drug exposure and resultant effect(s). As a consequence, drug development has been inefficient by relying on trial and error at the clinical stage, not to mention, extremely expensive. According to a 2013 investigation by Forbes, 95% of drug candidates failed to demonstrate either safety or efficacy in clinical trials. Moreover, in 2012, the average cost of bringing a drug to market was $4 billion.
Model-based drug development (MBDD) supports all relevant decisions with quantitative information. It employs mathematical models to build a conceptual framework for the drug, its competitors, the target disease, patient outcomes, and cost of treatment. MBDD can be applied across drug discovery, pre-clinical studies, and clinical trials.
Physiologically-based pharmacokinetics (PBPK) modeling—a type of biosimulation—is gaining attention from regulatory agencies, including the FDA, for its ability to predict drug disposition when information from in vitro systems are integrated into it (via a process called in vitro-in vivo extrapolation—IVIVE). It is often successful in predicting drug exposure levels based on patient and drug characteristics, concomitant medications, and more. PBPK can aid dose selection for various types of patients if the physiological and biological attributes of the population are known. In addition, it can anticipate potential drug-drug interactions (DDI). Models based on preclinical data can also be incorporated, with or without supporting clinical data. By providing a better understanding of the physiological mechanisms of drug PK, PBPK modeling and simulation has the potential to reduce drug toxicity, increase efficacy, and decrease the cost of bringing a drug to market.
The FDA rejects 16% of first-round submissions because of sub-optimal dosing recommendations. Certain populations (children, patients with rare diseases, pregnant women, etc.) present ethical or practical challenges to conducting clinical trials to determine dosing. PBPK M&S can be used to understand PK in these difficult-to-study groups. For example, pediatric dosing recommendations for an investigational drug were developed using a PBPK model that determined the dose for 12–18-year old patients that produced systemic exposure levels matching those in adults. This approach saved time and money on clinical trials and avoided exposing children to experimental medications.
The FDA estimates the cost of adverse drug reactions (ADRs) at $136 billion annually. DDIs are a significant subset of ADRs and are an important consideration for drug developers. DDIs can occur when a medication alters drug metabolizing enzymes that also metabolize concomitant medications. PBPK models incorporate information about how drug exposure changes with drug-induced enzymatic inhibition or induction. Thus, the models can predict and quantify the magnitude of potential DDIs and sometimes even eliminate the need for additional clinical studies.
In conclusion, PBPK methods use physiologically informed models incorporating relevant information from the study drug as well as concomitant drugs. Such models can also include other factors likely to affect exposure, such as disease state and severity, and factors associated with populations of interest: patient demographics, diet, relevant genomics, etc. This approach enables quantitative decision making from discovery through all stages of drug development.
Want to learn more about how PBPK modeling is changing the drug development landscape? Read our white paper, “Physiologically-based Modeling in Development Decisions and Regulatory Interactions.”