Regulatory Insights: Adaptive Clinical Trial Design – Overview and Critical Success Factors
The U.S. Food and Drug Administration’s (FDA) 2004 Critical Path Initiative,1 21st Century Cures Act of 20162 and subsequent issuance of guidance documents have driven increased interest in the application of adaptive clinical trial designs as sponsors try to bend the drug development cost and time continuum. Adaptive designs were implemented to accelerate clinical trial decision-making as part of the COVID-19 response under Operation Warp Speed.3 In this post, Kaitlyn Huff, associate director, and Patricia Forti Mann, senior director of U.S. regulatory solutions, provide an overview of regulatory considerations for adaptive design.
Adaptive design for drug and device studies
The pharmaceutical landscape is evolving as stakeholders pursue the use of real-world evidence, precision medicine, and complex innovative trial designs (CID) to increase efficiency of development. Innovative designs include master protocol and adaptive design trials wherein sponsors incorporate prospective modifications to one or more aspects of the design based upon accumulating data from subjects on the trial without compromising study integrity or validity.4
These methods are advantageous for both patients and developers, as demonstrated by their use in medical device and key therapeutic areas, such as oncology and infectious disease. A 2017 review of publicly available databases and the literature indicated a recent growth in publicly reported adaptive design use among researchers around the world.
As noted by Bothwell, Avorn, and Khan in a systematic review of adaptive design trials, the most frequently utilized trial adaptations were seamless Phase II/III (57%), group sequential (21%), biomarker adaptive (20%), and adaptive dose-finding designs (16%). Approximately one-third of trials used an independent Data Monitoring Committee, while 6% reported blinded interim analysis. Nine percent of the reviewed adaptive trials were used to support submission of a U.S. new drug application, and 12% were used for approval of a European Medicines Agency (EMA) marketing application authorization.5
Adaptive design relevance and utility are highlighted in the current COVID-19 environment as companies leverage innovative designs for products seeking to treat, prevent or mitigate the effect of SARS-CoV-2.
Included among the pre-planned changes an adaptive design may permit are:
- Refinement of the sample size
- Changes in the allocation ratio of patients to trial arms
- Identification of patients most likely to benefit to focus recruitment efforts
- Decisions to abandon treatments, doses or to halt the trial at an early stage because of success or lack of efficacy6
Ultimately, innovative designs help sponsors address development questions more efficiently7 and have a better chance of obtaining the relevant answer from a clinical trial. With appropriate adaptive design planning, a sponsor may implement changes to study design, conduct, and statistical hypotheses or analysis based upon prespecified interim analysis. Adaptive designs are especially useful in the pivotal stage if there are uncertainties about one or two aspects of the study.4 Should additional uncertainties exist, resolution of these questions prior to initiation of pivotal trials is advisable as this approach may be more efficient and increase the probability of successful trial outcomes.
As outlined in FDA guidance documents, a successful adaptive trial must:
- Control type-1 error
- Remove/limit statistical bias
- Employ prospective study design
- Maintain clinical trial integrity
FDA review divisions are not necessarily aligned with regard to incorporation of adaptive designs, and input may depend on the investigational drug product type, indication and patient population. Therefore, it is especially important to clearly define the advantages of the proposed design with regard to ethics, feasibility and efficiency. Clinical and/or nonclinical data should support dose justification during Phase III to mitigate any potential safety issues since FDA will not be privy to interim analyses. This is especially pertinent for seamless adaptive studies used to select the most likely efficacious dose.
Alignment on the horizon
Multinational firms typically conduct global trials, which means the FDA is not the only competent authority that needs to align with the clinical trial design. The International Council for Harmonisation Final Concept Paper E20: Adaptive Clinical Trials dated 7 November 2019 and endorsed by the Management Committee on 18 November 2019, recognized the ongoing challenge in harmonizing European and U.S. regulatory agency guidance on adaptive trial design and has committed to formation of an adaptive design working group to facilitate such guidance. It is anticipated that this effort may take from three to four years.8
Key potential advantages of adaptive studies include increased flexibility and efficiency in development. Preplanned changes to the study offer supplemental ethical practices as subject treatment is optimized throughout the trial by eliminating ineffective or unsafe treatment arms prior to the initially scheduled end of trial.4,9 Through these practices, sponsors may gain deeper understanding of the investigational product where an adaptation facilitates risk analysis at key study milestones. Thus, adaptive trials permit strategic deployment of resources in development of innovative therapies.
Adaptive designs are multifaceted and often require pretrial statistical simulations and specific analytical methods which necessitates additional effort during design. Early planning efforts should also mitigate logistical challenges that could impact the interpretation of clinical trial results. These are successfully managed by leading-edge companies that gain advantage by initiation of strategic risk assessments, early planning with a CRO partner with CID experience and facilitation of effective health authority meetings. PPD’s experience submitting adaptive protocols to U.S. and European regulators highlights the acceptability of such innovative trial designs. While CID require more time on the front-end, they can accelerate the overall development timeline and thus maximize cost savings.10
Critical factors to drive successful implementation of adaptively designed clinical trials
Thoughtful consideration and risk-analysis should be performed when taking advantage of the strategic opportunities afforded by adaptive clinical trial designs. The decision as to whether to proactively plan for two confirmatory randomized Phase III clinical trials or to implement one adaptive and one “traditional” Phase III study present both a challenge and an opportunity for many sponsors that will directly impact their overall drug development operational planning. Key considerations for successful implementation of innovative adaptive trials include the following:
- Timing of interactions with FDA and other national competent authorities, such as the EMA, to ensure alignment of Phase II/III study designs to support planned timeline for submission of the marketing authorization
- Allow adequate planning time upfront to develop robust statistical simulations that control the risk of false positives (type I error) and false negatives (type II error). FDA will ask about error during meetings to discuss confirmatory trials; be prepared. Ensure your explanations are understood by both the FDA statistician (more detailed) and clinician
- Statistical analysis plans should address both interim and final analyses
- Inclusion of an unblinded independent consulting statistician on the Data Monitoring Committee (DMC) to help support decision-making and to avoid the introduction of bias
Comprehensive planning, thoughtful execution and teams with strong adaptive design expertise can deliver game-changing advantages to sponsors resulting in successful trial execution. When appropriately implemented, adaptive studies present an exciting opportunity to bend the cost and time continuum of drug development, thus changing the current development paradigm.