Multiple Dynamic Endpoints and the Role of Endpoint Adjudication Committees in Clinical Trials
How do you know your experimental therapy is working? Sometimes, it can be as simple to determine as comparing drug A to placebo and measuring the physiological result with an approved assay. But what does “working” really mean when efficacy means something subjective, like quality of life, reduced psychiatric events, or less symptoms?
In most cases, the results you are really looking for and trying to prove in a trial can be complex. How do you strategically plan your therapy and research objectives to improve the chances you will satisfy what the regulators are looking for and ultimately get your therapy approved for marketing? Can you use surrogate endpoints, multiple endpoints, what statistical power will you need, will diversity play a factor in planning your research strategy? All key questions to consider and address before the therapy even leaves the lab.
One area receiving increased focus from the Food and Drug Administration (FDA) are trial designs incorporating multiple endpoints to support efficacy. FDA recently released guidance on how sponsors can approach multiple endpoints in their clinical trial design in a way which will likely satisfy the regulators when it comes time for marketing approval. This latest guidance also suggests and reenforces the need for an independent review of clinical events to establish unbiased verification of the endpoint, as outlined in the agency’s 2006 guidance. Similar guidance has been issued from the European Medicines Agency (EMA).
Why Issue New Guidance?
Therapies are increasingly complex – innovative cell and gene therapies, immunotherapies, and even traditional small molecule drugs are interacting with the human body in more advanced, multi-factored ways.
When trying to demonstrate a therapy’s efficacy with potentially multiple complex modes of interacting with the human body’s cellular functions, sponsors naturally will include multiple endpoints to highlight and maximize the positive impacts of the therapy.
In the new guidance, FDA alludes sponsors are increasingly including multiple endpoints in their trial designs presented to the agency. That the agency felt it was necessary to issue guidance on this specific topic implies there have been problems with multiple endpoint designs. These problems likely lead to delays in study startup and regulatory approval of innovative new therapies.
When utilizing multiple endpoints to demonstrate potential efficacy of a new therapy, the recognized challenge is the trial designs require increasingly complex statistical methods to account for multi-variant and confounding variables.
FDA is concerned these increased statistical complexities in a trial may increase the “likelihood of making false conclusions about a drug’s effects with respect to one or more of those endpoints … if there is no appropriate adjustment for multiplicity.” In short, FDA is acknowledging this is complex area, so make sure all clinical development and biostatistical experts are well-versed on how to design these types of trials to ensure startup and/or regulatory approval is not delayed.
Approaching Multiple, Complex Endpoint Designs
The task of developing protocols containing multiple endpoints is not impossible, but it needs to be appropriately thought out and integrated into the trial’s endpoint and statistical design.
Determining the appropriate endpoints, and how best to measure them, is the first step. Sponsors need to carefully consider how they will prove safety and efficacy in a manner sufficient to satisfy the regulators. FDA is concerned that “failure to account for multiplicity when there are several endpoints evaluated in a study can increase the chance of false conclusions regarding the effects of the drug.” With that in mind, sponsors incorporating multiple endpoints in a trial must rigorously evaluate how to account for the array of endpoints potentially interacting or interfering with each other in such a way as to mask the true safety and efficacy of the therapy.
In addition to having multiple primary endpoints, when evaluating the appropriate endpoint design for a trial, sponsors need to additionally consider the use of composite endpoints, co-primary endpoints, and multi-component endpoints. All of these depend on the nature of the disease condition and the modality of treatment interaction.
Endpoint designs are as unique and varied as the disease conditions and therapies themselves. Developing the endpoint design and overall approach of a clinical trial program is mostly science, but there is a bit of art to it as well.
Working with Statisticians
FDA’s guidance highlights a number of statistical designs potentially suitable to account for multiplicity, including:
- The Bonferroni Method
- The Holm Procedure
- The Hochberg Procedure
- Prospective Alpha Allocation Scheme (PAAS)
- The Fixed-Sequence Method
The sponsor’s biostatistical team and the independent data safety monitoring board (DSMB) statisticians should be well versed in these statistical methods. These groups are equipped to ensure the appropriate method is selected and detect safety issues within the data, so participants are appropriately protected.
Engaging an Independent Endpoint Adjudication Committee (EAC)
With complex and multivariant endpoints comes an increased need for medical experts to carefully scrutinize the clinical events and determine if the endpoint(s) have been met. Typically, an independently administered EAC handles this evaluation. Also known as a clinical event committee (CEC), EAC members evaluate a clinical event in the context of the protocol and the individual participant’s medical information. There are usually three or more reviewers who evaluate the dossier and determine whether the event meets the endpoint.
An EAC’s determinations are incorporated into the overall trial data. As experience and data is accumulated, the statistical model begins to be satisfied. Eventually, an evaluation by the independent DSMB of accumulated data is completed to verify there are no emerging safety or futility issues.
The Importance of an Independent EAC for Regulatory Approval
The core role of an EAC is to have medical experts independently evaluate a clinical event to determine if a defined endpoint or clinical threshold has been met, or if the event should be categorized in some way (e.g., caused by the study drug or by an underlying condition). Such work requires medical judgment and a lot of associated information to make the determination. This is particularly important when untangling the complexities of multivariant endpoints.
Both FDA and EMA stress in their guidances the importance of the EAC evaluators being independent of those sponsoring or conducting the research. Using an independent EAC provides assurance to the regulators the endpoint determinations have been carefully evaluated by multiple experts without the impression of bias.
What to do When Developing Clinical Trials
FDA is clearly concerned about this topic – thus the need to issue guidance. Sponsors should take note and ensure their trials utilizing multiple endpoint designs have appropriate statistical designs to account for multiplicity.
Study the statistical methods, plan the trials, and talk with experts and FDA early about trial design plans. Do not end up running a trial only to find the endpoint design, or the associated independent EAC/DSMB plans, are not adequate to satisfy the regulatory agency.
Bottom line: Designing a clinical trial approach and program is complex. You can do a lot of things in a trial; it’s mostly science and part art. FDA is watching this area of trial design, so if you are planning a trial with multiple endpoints, be certain you have the right clinical and statistical experts as part of your team.