Computational modeling and simulation (M&S) have been recognized as powerful tools in facilitating drug and medical device development and regulatory research, with the capability to accelerate access to safe and effective products. Since M&S was named a strategic priority at FDA almost a decade ago,1 its use by FDA and industry has grown. By 2017, almost all new drug applications contained some elements of M&S.2 To support its growing use, the FDA established a 5-year model-informed drug development (MIDD) pilot program the following year under the Prescription Drug User Fee Act VI to enable early discussions between regulators and sponsors on M&S strategy issues.3
The Agency sees a wide range of applications for M&S, some of which are considered mature, and others still emerging. For regulatory purposes, it’s important to identify best practices to apply M&S approaches and evaluate model “credibility,” that is, the predictive reliability of the model. As guidance on M&S emerges, we continue to see a considerable amount of flexibility, rather than a standardized approach, to evaluating model credibility and adopting terminology in describing elements of model assessment. However, a uniform approach may ultimately lead to more consistent and streamlined regulatory decision-making.
CDER Researchers Explore a Credibility Assessment Framework in the Context of Model-Informed Drug Development
To begin aligning approaches across FDA, regulators in the Center for Drug Evaluation and Research (CDER) considered an evidentiary framework that was developed by the American Society of Mechanical Engineers and colleagues in the Center for Devices and Radiological Health (CDRH), where it is currently being used in medical device development.4 This Verification and Validation standard, known as the V&V40, contains many best practices in model evaluation, but it doesn’t specify activities or criteria for particular modeling approaches or applications. Instead, its high-level structure provides a risk-based framework for determining how much and what type of evidence could effectively underpin a computational model for decision-making within a given context (e.g., development of a particular drug). Overall, this framework has the potential to minimize the risk of erroneous decisions based on computational predictions. The generality of the standard and its potential to de-risk the use of M&S led CDER to consider its utility in drug development.
In collaboration with colleagues in CDRH and the Center for Biologics Evaluation and Research, the CDER Office of Clinical Pharmacology evaluated the applicability of the V&V40 standard to physiologically based pharmacokinetic (PBPK) models, a type of computational modelling used across many Centers in evaluating regulatory submissions and conducting research. Applying principles of the V&V40 to case studies, the FDA researchers have reported that key concepts could be relevant to PBPK models for drug development and potentially expanded to other modeling approaches.5 Broader adoption of the framework could lead to a more consistent approach to computational model assessment, irrespective of the model type or intended application. This realization prompted discussions about harmonization with other CDER Offices as M&S programs evolve and grow. These discussions were further informed by engagement with industry and foreign regulators at FDA’s public workshop on best practices for PBPK.6
The potential utility of the V&V40 framework, as outlined above,5 was recognized by the CDER Office of Pharmaceutical Quality, Office of Translational Sciences, and Office of Generic Drugs in guidance to stakeholders seeking to utilize PBPK modeling in biopharmaceutics applications for oral drugs.7 The guidance states that sponsors should consider using a risk-based approach, a key element of the V&V40, along with several other important factors when implementing PBPK for biopharmaceutics. Similarly, to explore its utility across M&S approaches, the MIDD pilot program recently asked sponsors to apply elements of the V&V40 to meeting requests and packages,3 as these elements could help to facilitate alignment between sponsors and regulators early in drug development. The broader applicability of the V&V40 to in silico methods has also been discussed by the European Medicines Agency and others in a recent publication.8
Explorations of the V&V40 across the FDA and beyond suggest the framework may be a good starting point for discussions on harmonization of model assessment. Whether the V&V40 or some alternative overarching framework is ultimately adopted, the value of a consistent approach to M&S evaluation and implementation is clear.
Overview of the V&V40 risk-informed credibility assessment framework. (COU, context of use; V&V, Verification and Validation)
The Spotlight series presents generalized perspectives on ongoing research- and science-based activities within CDER. Spotlight articles should not be construed to represent FDA’s views or policies.
1 FDA. Strategic plan for regulatory science. Accessed September 16, 2020.
2 FDA. Speech to the Regulatory Affairs Professionals Society (RAPS) 2017 Regulatory Conference. Accessed September 16, 2020.
3 FDA. Model-informed drug development pilot program. Accessed September 16, 2020.
4 The American Society of Mechanical Engineers. Assessing Credibility of Computational Modeling through Verification and Validation: Application to Medical Devices. V&V 40. The American Society of Mechanical Engineers. 2018. Accessed September 16, 2020.
5 Kuemmel C, Yang Y, Zhang X, et al. Consideration of a credibility assessment framework in model-informed drug development: potential application to physiologically based pharmacokinetic modeling and simulation. CPT Pharmacometrics Syst Pharmacol. 2020;9(1):21-28. doi:10.1002/psp4.12479.
6 FDA Meeting. Development of best practices in physiologically based pharmacokinetic modeling to support clinical pharmacology regulatory decision-making. Nov 18, 2019. Accessed September 16, 2020.
7 FDA. The use of physiologically based pharmacokinetic analyses—biopharmaceutics applications for oral drug product development, manufacturing changes, and controls. Guidance for industry.
8 Viveconti M, Pappalardo F, Rodriguez B, et al. In silico trials: Verification, validation and uncertainty quantification of predictive models used in regulatory evaluation of biomedical products. Methods. 2020;S1046-2023(19)30245-2. doi:10.1016/j.ymeth.2020.01.011.