Stacking the odds for success: Choosing biomarkers, endpoints, and designs for mid- to late-phase prevention trials

ED05-02

Upon the discovery of a handful of promising cancer prevention agents, a critical step is to conduct clinical trials to further test the agents and identify the most promising ones for large-scale prevention trials. The mid- to late- phase prevention trials are critical to the success of developing prevention agents. A few strategies will be introduced to maximize the odds for success.
 (1) Choosing the proper biomarkers: An effective prevention agent must be able to reach the target tissue and exert its effect. Standard PK-PD markers for measuring the drug delivery and metabolism can be used as a first step to measure the drug effect, for example PGE₂ level for COX inhibitors. Other molecular markers such as Ki-67 for proliferation, apoptotic index for apoptosis, viral load for anti-viral vaccine (e.g., HPV vaccine), and PSA for prostate cancer prevention have been used. The use of molecular imaging technique for target agents also offers a promising future.
 (2) Choosing the proper study endpoints: One of the major challenges in prevention trials is the low incidence of cancer during the trial. Using cancer as the study endpoint is not feasible in the mid-phase development and should be reserved for large-scale prevention trials. On the other hand, the intraepithelial neoplasia (IEN) can be an ideal alternative. IEN is often a precursor of cancer and can be modulated, for example, oral leukoplakia for oral cancer, polyp for colon cancer, and ductal carcinoma in situ for breast cancer.
 (3) Choosing the proper study designs: Key components for a successful trial design include - identify a high-risk population; enrich the study population for targeted agents; devise the best randomization scheme, such as the outcome-based adaptive randomization; and incorporate interim analysis for early stopping due to futility or efficacy. The use of neo-adjuvant design for prevention trials is also very appealing.
 Examples and practical considerations will be given to illustrate the above points with the goal of designing efficient and flexible trials to screen out losers and send off winners for further development.