A prime FDA strategy for reversing this trend is to clarify agency expectations for preapproval research, starting with oncology products and treatments for diabetes and obesity. The objective is to publish a series of guidance documents on the design and execution of clinical trials to help sponsors structure product claims and use standardized approaches for evaluating efficacy. New guidances will address trial design issues for therapies that treat different types of cancer at different disease stages, based on discussions with scientists at workshops and advisory committee meetings that address:

• the implications for selection of control groups for clinical trials (for example, placebo controls or control groups receiving standard therapy).
• the statistical evaluations that should be used to determine requirements for trial size, scope, and enrollment.
• how disease variations in subpopulations (categorized, for example, by sex, ethnicity, and/or age group) may affect expected outcomes and drug responses.
• safety concerns raised by treatment of study participants for concomitant conditions, which may involve taking other drugs.
• effect of disease or condition to be treated on pharmacokinetics.
• manufacturing or formulation issues that may raise questions about product consistency throughout a clinical trial.
• impact of decisions to exclude certain patient subgroups from initial trials based on an assessment of risks and benefits.
• measures for monitoring trials to ensure prompt response to data indicating safety or efficacy concerns.
• definition of safety endpoints that would lead to trial suspension.
• requirements for optimum presentation of risks and potential benefits as part of obtaining informed consent.
• efforts to assess the appropriateness of selection criteria for trial participants and of recruitment goals as defined in study design.

Joint efforts
To further encourage development of new oncology therapies, FDA is expanding its joint projects with research institutions to assess new approaches to product development. In April, officials from FDA’s Oncology Drug Products Division held a workshop with members of the American Society of Clinical Oncology (ASCO) to discuss appropriate endpoints for clinical trial design, focusing on studies for treating lung cancer. At the annual ASCO meeting in June, McClellan and National Cancer Institute (NCI) director Andrew Von Eschenbach announced plans to expand collaborations designed to facilitate development of new cancer therapies by

• developing standard approaches for evaluating markers of clinical benefit (biomarkers) that may serve as surrogate endpoints in clinical trials.
• creating a cancer bioinformatics infrastructure to improve data collection, integration, and analysis for preclinical and clinical research.
• collaborating on efforts to discover protein markers in the blood that can be used to detect and monitor drug response. This may provide a model for initiatives in areas such as diagnostic imaging and molecular targeting.
• developing clinically meaningful endpoints for evaluating cancer chemoprevention agents.

FDA officials also are looking for joint opportunities to evaluate endpoints and research policies for other cutting-edge technologies, such as cell and gene therapies, pharmacogenomics, and novel drug delivery systems. For example, FDA and NCI plan to expand joint programs on proteomics and genomics to develop standards for the safety, purity, and potency of tumor vaccines. FDA also is working with the American Society of Gene Therapy to define the design of toxicology studies to support licensure of such innovative treatments.

McClellan also said at ASCO that FDA is changing its accelerated approval policy to speed more new competing cancer treatments to market. This policy permits FDA to reduce premarketing study requirements for important therapies that meet “unmet medical needs” for treating cancer, AIDS, and other critical diseases. In return, sponsors commit to conducting additional postapproval research to confirm initial results. However, the policy permits FDA to designate only one product in a treatment area for accelerated approval; sponsors of additional therapies for similar indications thus must conduct the usual range of premarket testing.

The policy has led to system “gaming” as sponsors try to define slightly different indications or patient populations to gain favorable accelerated treatment. Moreover, the only-one-accelerated-approval policy can delay patient access to possibly useful therapies. Under the new proposal, FDA would grant accelerated approval status to multiple agents that qualify, even if they treat the same indication. However, once one sponsor confirms clinical benefit in Phase 4 trials, no additional products would be granted accelerated status. The aim is to make it easier for new therapies to enter the market while also encouraging sponsors to complete Phase 4 studies as promised (see Keeping Commitments sidebar on delays in postapproval study completion).

New research pathways
While FDA scientists are looking to streamline and clarify policies, agency officials and the research community are exploring ways to improve the overall results of oncology research and development programs. Many cancer trials end up with weak and inconclusive data, often because pressure to enroll subjects quickly in new studies leads to underpowered and poorly designed trials. The FDA-ASCO workshop on clinical endpoints provided a forum to air concerns about whether cancer studies produce sufficient information to fully assess product safety and efficacy.

Another meeting in April, sponsored by Georgetown University’s Center for Drug Development Science (CDDS), addressed ways to overcome the shortcomings of traditional cancer trials. Cancer drugs have more Phase 3 study failures than most therapeutic areas, commented CDDS director (and former CDER director) Carl Peck. He and his colleagues attributed this to the research practice unique to cancer of using early clinical trials to identify a maximum tolerated dose instead of seeking an optimal dose, followed by underpowered, uncontrolled, proof-of-concept, Phase 2 studies. Consequently, many Phase 3 cancer trials reveal unnecessary toxicities or fail to confirm sufficient effectiveness.

CDER pharmacometrics team leader Joga Gobburu and other speakers emphasized the value of determining a dose that produces maximum effectiveness early in drug development. Gobburu noted that this approach helped Novartis develop Gleevec (imatinib mesylate) by identifying doses that minimize side effects early in clinical development, leading to more accurate and useful labeling. Workshop participants concluded that defining proof of concept and identifying best doses in randomized, controlled Phase 2 trials can improve the success rate of Phase 3 studies.

Seeking study subjects
Designing successful oncology trials may help sponsors deal with the continuous challenge of enrolling a sufficiently varied subject population in relatively small oncology trials. A study, by FDA staff members, of elderly cancer patients found that even though 60% of new cancer diagnoses involve older patients, few of them (32%) enroll in clinical trials.¹ Underrepresentation of people age 65 and older in oncology trials makes it difficult to determine optimal doses and treatments for older cancer patients.

Another FDA analysis indicates a notable increase in oncology studies performed outside the United States. In the last two years, about 75% of clinical trials submitted to FDA to support applications for anticancer agents used non–U.S. sites; in 1998–99, that figure was about 50%.² Similarly, the number of trials conducted exclusively in the United States has declined from a high of 80% to about 24% today. Most of the non–U.S. studies are being conducted in Europe and Israel, reflecting sponsor desire to register new therapies simultaneously in major markets. The trend toward global trials also may expedite the development of new therapies by increasing access to small patient populations, such as children.

References
1. Food and Drug Administration, elderly patients in trials for cancer drug registration: A 7-year experience by the Food and Drug Administration (FDA, Rockville, MD, 2003).

2. “The Pink Sheet,” 65 (024) 31 (June 16, 2003).