The Business Case for Modeling and Simulation (M&S) - FDA’s Evolving View of the Role of M&S in Drug Development and Approval
by Daniel Weiner, PhD
Clinical development costs continue to escalate, leading to a “constipated pipeline” where sponsors cannot afford to progress valuable drug candidates to late stage development. Inefficient decision-making and poor knowledge management are root causes of this wastefulness. Decisions are made without the necessary inputs, are not quantified, are focused on the wrong issues, and suffer from loss of knowledge when staff transfers. FDA acknowledged these problems in its publication last year of "Innovation and Stagnation: Challenge and Opportunity on the Critical Path to New Medical Products,1" which proposes utilization of model-based approaches to improve drug development knowledge management and decision-making.
FDA’s Evolving View
The move to increased reliance on model-based drug development at FDA is in fact several years old. One of the earliest indicators of FDA’s growing acceptance of modeling and simulation was FDA’s publication of "Guidance for Industry, Population Pharmacokinetics." In that document, FDA stated that the use of the population PK approach could “help increase the understanding of quantitative relationships among drug input patterns, patient characteristics, and drug disposition.” FDA also pointed out that population PK studies are often necessary to determine changes in kinetics in some populations for which it's not practical to run separate studies, for example drug interactions, gender, and age. Subsequent publications have affirmed FDA's growing confidence in model-based drug development, as indicated in "Guidance for Industry, Exposure-Response Relationships: Study Design, Data Analysis, and Regulatory Applications” and “Guidance for Industry, Investigators, and Reviewers: Exploratory IND Studies".
More recently, FDA has continued to incorporate model-based drug development into its review process with its new focus on End of Phase IIa meetings. The stated purpose of the End of Phase II meeting with sponsors is to reduce unnecessary failures in late stage clinical trials (Phase IIb, III) through nonbinding, scientific review of exposure-response data from preclinical and clinical trials. FDA expects preparation for this meeting to include a significant modeling and simulation component to analyze and integrate relevant data to identify optimal doses, inform future study designs, and assess dose adjustment strategies for special populations. Another stated objective of the End of Phase IIa meeting is exploration of newer topics of uncertainty, such as QT trial design, and integration of pharmacogenomic data.
From these interactions with FDA, sponsors can improve and update their models based on Phase IIb experience and to plan for Phase III trial design. The models can be used as an important knowledge management tool for the traditional End of Phase II meeting. Based on uniformly positive FDA-sponsor experience from a two-year pilot program2, a draft End of Phase IIa Guidance was issued earlier this year.
The case for the value of modeling and simulation is becoming increasingly well documented. A recent analysis by FDA reviewers of 244 NDAs showed that 42 used modeling to some degree3. Pharmacometric analyses were pivotal in regulatory decision making in over half of these 42 NDAs. Pharmacometric analyses were used to provide evidence of effectiveness, assess benefit risk, review targeted safety studies, develop approval criteria, and evaluate clinical applications of failed bioequivalence studies. They were also used to formulate dosing instructions, including selecting the dose and regimen, individualized in doses, evaluating dosing in special populations, assessing drug interactions, and describing the time course of effects. Finally, they were used to provide warnings and precautions on labeling.
Of the 14 reviews that were pivotal to approval related decisions, five identified the need for further trials while six reduced the burden of conducting additional trials. Importantly, pharmacometric analyses were recognized as being used in trial design to select dose or exposure ranges for registration trials, and derive optimal sampling schemes to assess exposure and response. In addition, FDA has used modeling and simulation to evaluate alternative primary analysis benefits, pivotal bioequivalence criteria, and to compare competing recommendations in FDA Guidances.
The Path Forward
The several Guidances, the new focus on End of Phase IIa meetings, and FDA-led applied research provide substantial evidence that FDA is moving in a direction of increasing reliance on quantitative modeling and simulation. FDA’s movement in this direction has begun to drive trial designs and involve FDA and sponsors in more intensive communication about modeling assumptions and model-based insights to support development decision-making.
For the full potential of modeling and simulation to be realized, investments are required in several key areas. First, investment is needed in modern technical infrastructure, including a fully integrated PK/PD data management system, collaborative software tools, automation of routine tasks, and adoption of standards. A second investment area is the commitment to model building as intellectual capital, with the development and updating of “standing” drug-disease, trial and market models as a basis for therapeutic area knowledge management. Finally, adoption requires training of technical personnel and organizational leaders to execute modeling and simulation work. Effective execution itself requires a combination of biomedical knowledge and sophisticated quantitative skills that include probability and statistics. Modeling and simulation project leaders must have the full commitment of management and be able to effectively communicate with a broad range of organizational stakeholders.
FDA’s clear message on the importance of modeling and simulation underscores the need identified in its “Critical Path” whitepaper for a better product development toolkit. Model-based drug development offers an interdisciplinary set of quantitative methods, predictive approaches and strategies to increase drug development productivity, improve clinical quality and commercial performance of new therapies, and provide greater organizational transparency for decision-making.
1http://www.fda.gov/oc/initiatives/criticalpath/whitepaper.html
2http://www.fda.gov/ohrms/dockets/ac/05/slides/2005-4194S1_08_Powell.ppt
3Bhattaram VA et al. Impact of Pharmacometrics on Drug Approval and Labeling Decisions: A Survey of 42 New Drug Applications. AAPS Journal. 2005; 7(3). Available at: http://www.aapsj.org/view.asp?art=aapsj070351