Adaptive Clinical Trials
How will they affect CTM providers?
by Kristin Brooks
Given the so-called "innovation stagnation" of drug compounds moving though the mire of the clinical trial process, both Pharma and FDA are mounting efforts to establish and promote adaptive clinical trial designs that could potentially improve trial outcomes. An adaptive trial design can be characterized as a trial that makes use of the data collected throughout the process to make decisions or modifications to that trial. This type of design should empower or optimize collected data, as well as improve patient benefit.
So why haven't these trials been used sooner? Scientific advances made in the last decade, particularly in the area of drug screening, now enable those involved in drug development to learn more about the safety and efficacy of new compounds earlier in the game, exposing fewer patients to drugs that don't work. The FDA's Scott Gottlieb, M.D., deputy commissioner for medical and scientific affairs, stated, "The ability to fail faster is an important advance in science. If new scientific tools and approaches could help us learn more about new medicines more quickly and earlier in the development process, FDA wants to make sure that regulatory requirements are recognizing these tools and providing the appropriate flexibility so sponsors can make use of them."
Naturally sponsors have been frustrated with the rigid standards of traditional trials that do not fully utilize these scientific advances. The idea is to allow a sponsor to spend less time and money figuring out that a new compound doesn't work or has too many side effects. Adaptive trial designs may be the answer, but there are many questions/concerns that need to be addressed before jumping on the adaptive bandwagon.
According to industry estimates, the cost of drug development can exceed $800 million and take more than 10 years for a drug to reach the market. And to make matters worse, as many as half of Phase III pharmaceutical compounds fail. As a result, Pharma and FDA have been forced to rethink traditional trial designs in an effort to increase the success rates of drugs. The rather limiting and inflexible traditional approach to drug development is fast becoming part of an industry makeover, where clinical trial design is concerned. The goal is to utilize innovative tools that generate improved scientific data.
Currently clinical trials are generally highly experimental; drugs are tested on general populations, seeking a clinical response and a treatment effect that is better than an existing therapy without creating a slew of new side effects. This approach is extremely focused. For example, decisions regarding sampling during a trial are made in advance and remain unchanged, with patients being split and assigned to one of two treatment options. Then, when the experiment is finished, a decision is made as to which treatment was more effective. The statistical information that results from this method is related to how large populations with the same or similar condition are likely to respond to a treatment.
Where a traditional trial can fall short in today's drug development realm is in the inflexibility of design and analysis in the trial, limiting the ability to incorporate scientific data that would help determine which new treatments have a better chance of benefiting subgroups of patients. Traditional trials are typically very large and patients are unnecessarily exposed to inferior therapies. "Randomized double-blind, placebo controlled studies have traditionally been the gold standard in trial design. It's basically an approach where we make our best assumptions, design a trial based on those assumptions and conduct the trial. At the end we look at the analysis and make interpretations," stated Dr. Dominic Labriola, vice president of global biometric sciences at Bristol-Myers Squibb.
What Is An Adaptive Trial?
Despite the numerous types of adaptive trials under investigation, there seems to be a consensus in the industry as to the definition of what an adaptive trial is. Both industry and FDA have been addressing this "hot" topic during the last six months.
In a nutshell, adaptive trial designs allow the flexibility to review results of a trial on an interim basis and modifications to the trial design may be made in order to improve the chances of a positive result, or stop the study if it appears ineffective. This has the potential to save time by preventing the number of study failures, and to reduce the number of subjects required. Some of the more common design protocols being evaluated by the FDA include an adaptive dose-ranging model and 'seamless adaptive' trial designs.
The dose-ranging model modifies the randomization of a trial to refocus patients to the dose (or doses) that appear to be most beneficial or that have the greatest benefit with regards to safety and efficacy. Therefore, as a study grows, a particular dose may be eliminated because it looks too toxic or has too many side affects. The adaptive dose-ranging design is considered a very reasonable type of trial design to use, according to Dr. Labriola. With this design, ideally you minimize toxicity while data is obtained on the maximum tolerated dose.
Where you need to maximize the value of subjects or create efficiencies, apparently 'seamless adaptive' trial designs are the way to go. The adaptive seamless design combines two traditional trials, in particular Phase II and Phase III, into a single trial, which can reduce cost and time to market. By combining the goals of these two phases, sponsors wouldn't have to start a Phase III trial from scratch, as the data is continuous. Successful Phase II trials and unsuccessful Phase III trials are a cause for industry concern and this adaptive design has the potential to alleviate some of the pitfalls in-between phases. In particular when dealing with a hard-to-find patient population or a disease that is not very prevalent, seamless designs allow sponsors to get the most out of the subjects involved in the trial.
In a speech given at the 2006 Conference on Adaptive Trial Design, Dr. Gottlieb of the FDA said, "The traditional, highly empiric statistical approach has had the dominant, and often exclusive, role in drug development. A side effect of this approach is inflexibility, which in turn limits innovation in the design and analysis of clinical trials. There are potentially better alternatives, by enabling more trials to be adapted based on knowledge that can help predict whether patients will respond to a medicine." Dr. Gottlieb emphasized that these approaches to clinical trials can help us learn more about safety and benefits of drugs, in potentially shorter time frames, exposing fewer patients to experimental treatments. The result of which is more efficient clinical trials that are more appealing with respect to patient enrollment.
The FDA is taking steps to facilitate the continued development of adaptive clinical trials, and according to Dr. Gottlieb, "the opportunity couldn't be riper." Of the guidance documents currently being drafted by the FDA—as many as five in all—one will guide sponsors on how to look at multiple endpoints in the same trial. This guidance is expected to be complete as soon as January of 2007. Another guidance deals with enrichment designs that can help increase the power of a trial to detect a treatment effect, potentially with fewer subjects. The other three documents, which the FDA said would take longer to draft, will cover non-inferiority trial design, adaptive designs, and one on how to deal with missing clinical trial data.
While adaptive designs can create efficiencies in the drug development process, there are risks involved. Adaptive trials are certainly more complicated to design and analyze. Current industry concerns relate to statistical evidence; operational logistics and data collection; trial integrity and managing the privacy issues that arise.
So, why aren't we seeing more use of adaptive designs? According to Dr. Mark Cheng, technical director of research biostatistics at Millennium Pharmaceuticals, "Statistical methods are available for the most common adaptive designs, but for more complicated trial designs, the methodologies are still in development. Furthermore, without regulatory guidance, naturally there are fears that the protocol with this innovative approach may be rejected and cause delay."
Also, the industry needs to figure out ways to appropriately control the error rate associated with adaptive trial designs. According to Dr. Labriola, "First and foremost we have to control the potential for a false positive or type I error rate, which is basically the probability of making the wrong decision based on examining data on a periodic basis. You're potentially increasing your chance of making a wrong decision with repeated looks."
Operationally, adaptive trials require a more collaborative working environment and place a lot of burden on internal and external data monitoring with respect to trial modifications. Dr. Cheng added, "When dealing with real-time data collections and analyses, data monitors must be able to rapidly integrate knowledge and experiences from different disciplines into the decision making process."
Dr. Robert O'Neill, director, office of biostatistics at FDA, warns that adaptive designs are not without risk: "It's not a free lunch; there are risks in carrying out such designs." For example, unblinding a trial can cause bias and jeopardize the integrity of the trial. With traditional trial designs, data is kept under lock and key from all parties involved until the trial is complete. FDA recently issued guidance on the conduct of data monitoring committees in clinical trials and that guidance contains a very strong message with respect to confidentiality for unblinded interim analysis and the sharing of results outside of the data monitoring operation and the risks in not maintaining that secrecy. "An overarching theme here is that adaptive designs are not an excuse for poor planning or after-the-fact adaptations to salvage studies. While there's a lot of interest in these trial designs, my advice is that the industry should be relatively critical and cautious about the best circumstances under which adaptive designs might be most appropriate," Dr. O'Neill continued.
An "Adaptive" Service Industry
The FDA and sponsors are on the same path to enlightenment where adaptive clinical trials are concerned. Although still early in its evolution, there is much anticipation to use these trial designs in the near future. So, what is the prospective impact of these trial designs should they become widely adopted? Adaptive clinical trials could potentially impact all organizations involved in carrying out clinical trials, including those that supply clinical trial materials.
Dr. Cheng at Millennium remarked, "For a classic design, the amount of material required is fixed and can be easily planned before the trial starts. However, for an adaptive trial, the exact amount of materials needed is not clear until a later stage. Also, the next dosage for a site may not be fully determined until the time of randomization. Therefore, the vendors may need to develop a better drug distribution strategy using so-called operation theory or simulations." He also implied that the drug manufacturers that can provide the materials adaptively would have a better chance of success. "Depending on the types of adaptive design, there might be requirements in packaging and shipment to be more flexible and quick. It may also require quick and accurate efficacy and safety readouts. The electronic drug packages with an advanced built-in recording system would be helpful," he continued.
From the regulatory side of things, Dr. O'Neill added, "With the outsourcing of clinical trials these days to CROs, a lot of serious discussions are going to need to take place regarding the SOPs [standard operating procedures] for carrying out these trials so that the integrity of the trial is maintained. There's a lot of infrastructure that has to be discussed and set up in advance."
Ineffective clinical development strategies have the industry seeking new ways to improve the drug development process. For companies looking to produce new life-saving medications and recover colossal investments, the current development process is daunting.
Although still in the early stages, adaptive designs have captured the interest of many in the industry and, according to the FDA, are becoming a more common element in SPAs—agreements between FDA and sponsors in the early development process mapping out requirements for later trials. Clinical trial design is undergoing change with the potential to greatly impact the industry. With sponsors and FDA paving the way for adaptive trials, these designs raise questions as to how this type of trial could affect clinical trial providers, namely CMOs with logistics concerns and CROs with operational concerns.
We can expect to hear a lot more about adaptive designs once the FDA irons out some of the guidelines, as they are expected to do early next year.