What is Model-Informed Drug Development?

Model-Informed Drug Development (MIDD), sometimes referred to as Model-Based Drug Development (MBDD), is a process whereby key program decisions are supported by mathematical models and simulations that predict the likelihood of success for the drug. By using mathematical and statistical methods, one is able to build models of drug concentrations (pharmacokinetic) and/or drug responses (pharmacodynamic) over a time course. Such models allow one to understand how various dosing choices (e.g., dose amount, frequency, and duration) can affect drug concentrations while also demonstrating the relationship between drug concentration and the desired or undesired pharmacodynamic responses. In addition, these models help to characterize the PK/PD variability of drugs and assist in understanding the clinically relevant factors contributing to variability.

What can MBDD do for my Program?

Developing a drug is expensive and becomes progressively more expensive at later stages of development. MBDD may be able to help you save large amounts of time, money, and other resources. Even before your drug reaches the clinic, nonclinical modeling and simulation approaches can be used to predict human PK and establish an appropriate starting dose. MBDD can be especially useful if you already have some clinical data for your drug. Existing data can be used to predict important human responses and streamline development. For example, MBDD can be used to:

• Predict drug PK in different populations
• Avoid a full TQT study by using a cQT Model (assuming sufficient data exist from other studies)
• Justify doses and predict responses for multiple dose administration (assuming data exist from a single dose study)

MBDD Applications at Various Stages of Drug Development

MBDD can play a critical role throughout the drug development process, including: 

• Drug Candidate Selection (Pre-IND): A model-informed approach can utilize available in vitro and/or in vivo data to predict the pharmacokinetic profile of a drug in humans prior to the first human exposure. These early predictions can support the rationale for selecting the first dose to administer to humans, including an acceptable safety margin relative to exposures achieved in nonclinical toxicology studies. 
• Early Clinical Development (Phase 1): Early human pharmacokinetic (PK) or pharmacodynamic (PD) data can be used to develop the next stage of models of human exposure and/or pharmacodynamic response. A real-time, model-informed approach may be particularly useful to guide dose escalation during the conduct of ascending-dose studies. Upon completion of these studies, simulations based on the final model(s) can be a valuable resource when designing and optimizing longer-term studies. 
• Proof-of-Concept (Phase 2): Data collected at the proof-of-concept stage can be used to develop even more robust models. At this stage of development, model-informed predictions can be indispensable for selecting the optimal study design, doses, and dose regimens to progress into Phase 3. 
• Phase 3: During late-stage clinical development, PK and PD data are typically collected from a broad sample of the target population. These data allow further development of PK and PD models in preparation for regulatory filing and marketing. A key use of MBDD at this stage is for characterizing the variability in drug concentrations and drug response. Identification of demographic factors (e.g., age, body weight, renal function, hepatic function) that impact variability is often a critical step in the development of these models. Information gleaned from these models often serves as a foundation for developing dosing guidance for special populations (renal impairment, hepatic impairment, elderly, children) and in consideration of other clinically relevant factors identified in the model.

Conclusions

Model-Informed Drug Development offers the benefit of more streamlined drug development along with the potential for considerably lower resource costs and a reduced time to market. MBDD is useful at any stage of development but is most beneficial when it is integrated early and continuously to help guide objectives and effective program decisions.