The Strategy, Tactics, And Process Behind Successful Seamless And Adaptively Designed Clinical Trials

By Pol Boudes, M.D. and Harold Shlevin, Ph.D.

After months of planning – and working with the FDA – Galectin Therapeutics recently launched a seamless and adaptive design Phase 2b/3 clinical trial of belapectin in patients with nonalcoholic steatohepatitis (NASH) cirrhosis.  For those accustomed to just one type of clinical trial, you may ask, what is a seamless and adaptively designed clinical trial? So, here are some of the key takeaways we’ve learned in the process.

What Is A Seamless And Adaptive Design Clinical Trial?

Seamless means this trial will transition from phase 2 to phase 3 without interruption. An adaptive clinical trial evaluates a treatment by observing participant outcomes on a planned schedule and allows modification to the trial protocol in accordance with those observations. This way clinical trials are more flexible, efficient, and fast, enabling sponsors and investigators to make changes based on accumulating information, without needing changes in the protocol that require additional ethical or regulatory approvals.

Modifications may apply to treatment dose, patient numbers, patient selection criteria, and even the general conduct of a study (for instance, stopping for overwhelming efficacy or for a safety concern). In our case, the Phase 2 portion will focus on determining the right dosage to use in the Phase 3 arm, the appropriateness of the sample size calculation, as well as the risk/benefit of the belapectin. This will be achieved with a pre-specified Interim Analysis at the end of the phase 2 portion.

Since adaptively designed clinical trials were first introduced, they have been strongly supported by the FDA, the EMA and other health authorities around the world. A study conducted recently found that adaptations, such as seamless Phase 2/3 studies and adaptive dose-finding designs used in the belapectin trial, were common.

Advantages Of Adaptive Clinical Trials

There are numerous advantages to using an adaptive design in planning and conducting a clinical trial:

It saves time. Instead of doing a series of trials to test a drug candidate – with each trial taking time to set up and receive approval – sponsors can outline a single trial with multiple segments. The approach can demonstrate a drug’s efficacy and safety more quickly.

It saves money. Since it's a single trial, sponsors do not have to finance front-loaded costs involved in initiating multiple trials. Because an adaptive trial is one continuous trial, these startup costs are incurred only once. The use of adaptive designs – such as early study terminations due to futility or sample size re-estimation – have been projected to save sponsor organizations as much as $100 to $200 million annually.

The Interim Analysis also provides sponsors an early opportunity to determine if the trial is an overwhelming success, that its objectives might not be reached, or even that a risk/benefit concern has been raised. The trial could be stopped in these cases, eliminating the unnecessary costs of continuing the trial and providing the best and fastest decision for patients participating in the trial.

It is more flexible. Trial sponsors can react to events because scenarios have been planned in advance. There is no need to file protocol amendments if the changes are within the scope of the original trial plan.

Credibility of results are maintained, no matter how many changes are made, especially with blinded data or firewalls, where only a limited number of people have access to the results.

Trial subjects are used more efficiently. Patients can continue in the trial as it shifts from one phase to another, maintaining their treatment regimen and avoiding the necessity of recruiting another cohort of patients for a new trial. The ability to change dosing and even randomization means that fewer subjects are given ineffective placebo compounds, ineffective doses, or doses that are unnecessarily high. For a trial such as ours, where the progression of NASH cirrhosis is measured in months and years, an Adaptive Design is ideally suited to our need to maintain patients on the treatment over an extended period.

Disadvantages Of Adaptive Clinical Trials

It is important to note that adaptive design in drug development comes with some caveats:

More effort may be required, particularly in the planning phase. Therefore, the planning may take more time as more parties need to be consulted to reach a design that is accepted by all stakeholders. The belapectin trial is innovative in the sense that such a trial has never been done before, requiring significant discussions upfront, notably with health authorities.

It adds logistical challenges. The transition from one treatment arm to another must be carefully planned to maintain the integrity of the ultimate trial results

Data collection and analysis must be timely to aid decision making during the trial. The Interim Analysis depends on timely and high-quality interim data.

The statistical analysis is more complex. Adaptive designs require specific analytical methods to avoid increasing the chance of erroneous conclusions and introducing bias in estimates. It is especially difficult to control for type 1 errors in adaptive trials, otherwise known as false positives. The use of the more sophisticated statistical analyses is required.

Maintaining data validity can be challenging. If the data is unblinded during the Interim Analysis, this may jeopardize the credibility of the study.

Table 1. The differences between conventional trials and those using adaptive designs.

Feature	Conventional trial	Adaptive design
Design	Rigid	Flexible
Treatment arms	Two or three at most	Multiple
Hypothesis	Test the hypothesis	Gather data before hypothesis
Modifications	Protocol amendment required	Pre-specified changes allowed
Statistical analysis	Routine methods	Complicated Bayesian approach
Interim Analysis	Seldom performed	Done routinely
Role of Independent Data Monitoring Committee	Does little until trial phase is over	Active throughout the trial

Strategy, Tactics, And Process

Careful planning is needed in order to successfully execute an adaptive design in a clinical trial. Here are some considerations for the right strategy, tactics, and process that will increase the chances of success.

Strategy

An adaptive design isn’t right for all trials. Adaptive design trials are best suited for indications where not everything is known upfront. If there are open questions about the right dosage, patient population, event rate, or other criteria, an adaptive design may allow enough flexibility to resolve those issues while continuing to move the drug development process forward in a timely manner.

For more narrowly defined situations, where the endpoints are clear and they are well accepted, such as measuring bacterial blood culture, it may not be necessary to use an adaptive design. But in a new situation, the adaptive design approach allows sponsors to make changes that, taking a more traditional approach, would require an entirely new trial. Our earlier belapectin Phase 2 clinical trials hadn’t yet definitively resolved the dosage question, and doing a seamless Phase 2/3 trial allows us to determine the correct dosage and move without interruption into the Phase 3 part of the study.

It takes more time and thought up-front. Because all potential changes must be prespecified in the trial protocol, it is important to spend time to thoroughly understand any possible situations that might arise and plan out all potential changes made in response. Consulting with key opinion leaders and other outside experts in the disease and health authorities is strongly recommended.

Another area to be carefully planned for is the use of Interim Analysis. This may be a necessary step to determine what adjustments are needed in the trial, as each one carries a substantial cost in terms of the need for additional patients to establish statistical validity. The goal is to achieve a balance between the needs for assessing accumulating information and the risk of damaging the integrity of the trial. Be sure to minimize the number of Interim Assessments and keep the data blind whenever possible.

Think of the trial as a whole. The trial will have clearly defined endpoints, so any part of the study should advance the understanding of what it will take to achieve those endpoints. If, for example, the trial is a seamless Phase 2/3 study, what does the Phase 2 part of the study need to clarify before preceding to and ostensibly improving the chance of a successful Phase 3?

Carefully define what adjustments will be made. In an adaptive design trial, sponsors can make adjustments to the trial, provided they pre-specify what the nature of those are going to be. Making adjustments that aren’t pre specified would defeat the purpose of an adaptive trial.

Tactics

Streamline the trial. Clinical trials often include ancillary tests – such as biopsies or special cohorts – to explore interesting scientific questions. These efforts may advance scientific understanding of the disease, but they may not contribute to the primary goals of the trial, which is to prove the safety and effectiveness of a particular therapy. The process of running an adaptive design trial is complex enough that it is recommended to avoid exploring such side issues. For our own study, we eliminated the need for liver biopsies, both because they are not a primary outcome and because it is easier and more convenient for patients. We may lose some information, but the study will now be very much focused on clinical endpoints that are relevant in medical practice (e.g. esophageal endoscopy) rather than scientifically interesting ones.

Plan for patients to continue as the study shifts to the next part. For example, in a multi-phase trial, how can you optimize the number of patients participating in the Phase 2 arm that continue on in the Phase 3 study without jeopardizing the trial results? This not only provides benefit to the patients who continue with their treatment uninterrupted, but it also provides more safety information over a longer period, something that is clearly important to evaluate the risk/benefit of a drug.

Build in break points. The Interim Analysis provides an opportunity to end the trial early, if needed, and thereby eliminates the need to spend money or impose un-necessary constraints on patients on a trial that has already proven to be a success or a failure.

Process

Establish a Data Safety Monitoring Committee overseeing the study, independent from the sponsor, and a sponsor Trial Steering Committee, independent from the study team monitoring the trial. The data safety monitoring board for the trial will have access to the blinded and unblinded data. During the Interim Analysis, the board will make recommendations to the sponsor based on unblinded data. Those recommendations are given to a trial steering committee instituted within the sponsor, comprising staff members and experts who aren’t involved in the day-to-day operation of the trial. Steering committee members cannot be directly involved with the trial, otherwise the validity of the trial data may be compromised.

Beware of the statistical hit. Consider a traditional trial designed for 80% statistical power, where 300 patients are needed to reach that objective. The Interim Analysis in an adaptive designed trial would incur a statistical penalty for looking at the data. The magnitude of that penalty depends on how the examination is done– the rules are very complex – but now, instead of needing 300 patients, 400 patients might be needed to maintain the same 80% statistical power. Without those additional patients, the statistical power of the study could be 10% to 15% less, depending on the calculations, and that might not be acceptable for approval.

Using adaptive design is a trade off, and how the data is used in the Interim Analysis determines the magnitude of that trade off.

In The End

In principle, adaptive design should save time and money, but it also adds complexity that needs to be carefully managed. Sponsors should carefully consider the advantages and disadvantages that an adaptive design approach might bring to a clinical trial. The approach is not right for all trials, but it can be a good strategy to use in studies with unresolved questions. With careful planning, an adaptive approach can help avoid the need for additional trials and can shorten the time needed to prove the effectiveness and safety of a drug candidate. Galectin Therapeutics is likely to save years of time and effort during our Phase 2b/3 clinical trial.

About the Authors

Pol Boudes, M.D. is chief medical officer and Harold Shlevin, Ph.D. is president and CEO of Galectin Therapeutics, a developer of therapeutics that target galectin proteins. Galectin Therapeutics recently launched its NASH-RX seamless adaptively designed Phase 2b/3 trial of its galectin-3 inhibitor, belapectin, in patients with nonalcoholic steatohepatitis (NASH) cirrhosis.