Adaptive Clinical Trials: US FDA Offers More Advice On Bayesian Designs

Final guidance provides more details on clinical trial designs that use Bayesian adaptive features; says monitoring committee recommendations may deviate from the anticipated algorithm.

The US Food and Drug Administration's final guidance on adaptive clinical trial designs includes additional details on Bayesian design features and more flexibility in trial planning.

The guidance, Adaptive Designs for Clinical Trials of Drugs and Biologics, issued on 29 November, is nearly identical to the draft guidance published last year. However, at the request of industry, the agency expanded the section on Bayesian designs and made other tweaks to the document.

The draft document issued in September 2018 replaced a 2010 draft guidance. It provided examples of clinical trials with adaptive designs and more details on the documentation needed for FDA evaluation of the adaptive design and completed trial. (Also see "Adaptive Clinical Trial Designs: US FDA Provides Checklist To Begin Study" - Pink Sheet, 30 Sep, 2018.)

In comments on the new draft guidance, the Pharmaceutical Research and Manufacturers of America recommended that FDA provide additional guidance on the use of Bayesian adaptive designs, such as where the agency thinks the principles of Bayesian design approaches differ from other types of adaptive designs.

The final guidance states that in general, the same principles apply to Bayesian adaptive designs as to adaptive designs without Bayesian feature. The agency added: "Trial designs that use Bayesian adaptive features may rely on frequentist or Bayesian inferential procedures to support conclusions of drug effectiveness. Frequentist inference is characterized by hypothesis tests performed with known power and Type I error probabilities and is often used along with Bayesian computational techniques that rely on non-informative prior distributions."

"Bayesian inference is characterized by drawing conclusions based directly on posterior probabilities that a drug is effective and has important differences from frequentist inference," the guidance states. "For trials that use Bayesian inference with informative prior distributions, such as trials that explicitly borrow external information, Bayesian statistical properties are more informative than Type I error probability."

The agency notes that its draft guidance "Interacting with the FDA on Complex Innovative Clinical Trial Designs," issued in September, provides direction on what information should be submitted to FDA to facilitate the review of trial design proposals that use Bayesian inference.

The guidance also says that one common feature of many Bayesian adaptive designs is the use of simulations to estimate trial operating characteristics. Because many Bayesian methods themselves rely on extensive computations, "trial simulations can be particularly resource-intensive for Bayesian adaptive designs," it states.

Adaptations May Deviate From Anticipated Algorithm

The guidance defines adaptive design as a clinical trial design that allows for prospectively planned modifications to one or more aspects of the design based on accumulating data from subjects in the trial.

PhRMA also recommended that FDA revise the draft guidance to support greater flexibility related to the design, review, and acceptance of adaptive clinical trials, and the nature of allowable adaptations. For example, it said terms used in the guidance seem to suggest that adaptations should always follow a specific rigid numerical algorithm. PhRMA asked that the guidance make clear this is not required.

FDA added a paragraph that addresses this. "Although we recommend prespecification of the rules governing adaptations, monitoring committee recommendations might occasionally deviate from the anticipated algorithm based on the totality of the data," the guidance states. "If this type of flexibility is desired, the prespecified plan should acknowledge the possibility of deviations from the anticipated algorithm, outline factors that may lead to such deviations, and propose testing and estimation methods that do not rely on strict adherence to the algorithm."

The agency said that when completely unforeseen circumstances arise, it recommends that the sponsor address any potential design changes with FDA as soon as possible.

Non-Comparative Analysis

FDA added other clarifying sentences to the final guidance. In discussing the concept of non-comparative analysis, the document says, "it is possible to include adaptations based on a non-comparative analysis even in open-label trials, but ensuring that the adaptions are completely unaffected by knowledge of comparative data presents additional challenges."

The agency also slightly modified the section on adaptations based on a potential surrogate or intermediate endpoint. The guidance notes that most adaptive designs rely on ongoing monitoring of the primary endpoint or endpoints. However, in cases where a potential surrogate or intermediate endpoint exists that is correlated with the primary endpoint, and the primary endpoint itself is difficult or slow to ascertain, an adaptive design can be based on the potential surrogate or intermediate endpoint.

An adaptive design "could be based on a 2-month measurement of patient symptoms when the primary endpoint is the assessment of the same symptom outcome at six months," the final document states. The draft had specified a 3-month measurement and symptom outcome at one year.

Rejected Recommendations

FDA did not make several changes requested by industry. For example, PhRMA and the Biotechnology Innovation Organization had recommended that FDA revise the definition of "interim analysis" to be consistent with the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) E9, "Statistical Principles for Clinical Trials."

FDA retained the definition in the draft guidance that an interim analysis "is any examination of data obtained from subjects in a trial while that trial is ongoing and is not restricted to cases in which there are formal between-group comparisons." It specifies that the observed data in the interim analysis can include one or more types, such as baseline data, safety outcome data, pharmacokinetic, pharmacodynamic or other biomarker data, or efficacy outcome data.

Vertex Pharmaceuticals Inc. requested that the agency provide greater clarity regarding the regulatory path prior to submission of a special protocol assessment (SPA) for a trial with a complex adaptive design, and whether particular aspects of the study would be expected to be agreed upon prior to submission.

The agency did not provide further information in the section on interactions with FDA. Like the draft document, the final guidance recommends that SPAs for trials with complex adaptive designs only be submitted if there has been extensive previous discussion between FDA and the sponsor regarding the proposed trial and design.

Comments on the final guidance are due by 2 January.