Patricia F. Fitzpatrick Dimond Ph.D. Technical Editor of Clinical OMICs President of BioInsight Communications
A better early-phase clinical plan could address high attrition rates in Phase III.
With the highest failure rates of all therapeutic areas, cancer drugs cost their developers more money and consume more development time. Only about 8% of oncology drugs that complete Phase I testing ultimately achieve FDA approval.
The failure of Phase II trial results to translate into positive Phase III data constitutes a major and costly stumbling block in cancer drug development. Companies sometimes undertake Phase II without adequately understanding the drug, its pharmacology, and the condition they would like to target. The result can be a poorly designed clinical development plan unlikely to lead to the drug’s approval. Lack of clinical experience with drugs that have novel mechanisms of action contributes to trial failures and complicates the Phase II to Phase III transition.
Recent Phase III cancer drug failures underscore these issues. In January sanofi-aventis reported that a Phase III trial evaluating iniparib in patients with newly diagnosed metastatic triple-negative breast cancer failed to meet the co-primary endpoints of overall survival and progression-free survival. The Phase II study with iniparib in combination with gemcitabine/carboplatin, however, resulted in an improvement in overall survival and a high rate of clinical response in women with relapsed or refractory disease.
Another illustration that good Phase II results do not predict Phase III success comes from Antisoma and Novartis. Development of ASA404 (vadimezan) was halted after interim Phase III trial data showed that it was unlikely to meet the primary endpoint of overall survival when used in combination with chemotherapy for the second-line treatment of NSCLC patients.
The cost of Phase III failures adds up, particularly when a company spends money to acquire rights to a drug or buy a company outright for that drug. The vadimezan Phase III failure necessitated an intangible asset impairment charge of approximately $120 million in the fourth quarter of 2010 for Novartis’ pharmaceuticals division. Novartis paid Antisoma $75 million up front for rights to ASA404 in 2007 and a further $25 million upon initiation of the Phase III study. It also agreed to pay milestones of up to $980 million.
Design Criteria Subject to Debate
The idea that Phase II trial design may be an underlying contributor to Phase III drug failures is not new. Robert Glassman, M.D., Merrill Lynch managing director, succinctly stated the problem in 2007: “There is a lack of predictability in the way Phase II trials are conducted in cancer because few are done in a comparative fashion.”
As a result, he said, “problems arise in late-stage trials, which lead to high attrition rates. Noncomparative or uncontrolled trials have limited generalizability due to inherent biases based on trial effects. These trial effects include selection biases based on differential patient selection, investigator- and institution-based biases, over-enthusiasm for new agents—investigators have reasons to want to show that the new agent or combination under their study is superior to standard of care—and random and chance improvements from small numbers of patients in trials.”
In an effort to provide some guidance, the Clinical Trial Design Task Force (CTD-TF) of the National Cancer Institute (NCI) Investigational Drug Steering Committee (IDSC) published a series of discussion papers in the in March 2010 issue of Clinical Cancer Research. The IDSC developed formal recommendations about aspects of Phase II trial design “that are the subject of frequent debate,” including endpoints (response versus progression-free survival), randomization (single-arm designs versus randomization), inclusion of biomarkers, biomarker-based patient enrichment strategies, and statistical design (e.g., two-stage designs versus multiple-group adaptive designs).
In general, the IDSC “encourage[s] the use of progression-free survival as the primary endpoint, randomization, inclusion of biomarkers, and incorporation of newer designs,” while acknowledging that “objective response as an endpoint and single-arm designs remain relevant in certain situations.”
Using Adaptive Design
CROs that design and manage clinical trials for pharma and biotech companies believe strongly in the power of a well-designed and executed Phase II trial as part of an overall clinical development plan. “The relatively recent explosion in understanding the molecular pathophysiology of cancer and the identification of new targets is revolutionizing the way we conduct Phase IIa/IIb clinical trials in oncology,” Parexel’s Denis R. Miller, M.D., therapeutic area team leader, oncology/hematology, told GEN.
“The increased cost of clinical research, the intensely completive environment, the limited number of patients available and willing to enroll in clinical trials, the pressures of time and economics on physicians who are engaged in clinical trials, and the identification of validated biomarkers that accurately predict response are major and compelling considerations that that must be incorporated into the design and conduct of early-phase clinical trials,” Dr. Miller noted.
He encourages the use of adaptive designs, enrichment (enrolling only patients whose tumors express the drug target), randomized designs, and selection of endpoints that are surrogates of overall survival. The incorporation of adaptive designs may take a number of forms including changes in eligibility criteria, planned sample size, test statistic or analytic methods, hypothesis, primary endpoint, choice of dose groups or treatment arms, allocation to treatment to achieve balance or assigning fewer subjects to inferior treatment based on interim analysis, and use of an enriched patient population only.
Seamless two-stage designs have been proposed in which Phase I determines safety only via maximal tolerated doses. Phase Ib/IIa would refine that dose by evaluating efficacy and safety to determine the best dose.
Dr. Miller noted that the use of adaptive designs “should go a long way in increasing the likelihood of a successful early-phase clinical trial.” These novel approaches should deliver us more efficiently to the next step of conducting a large-scale, pivotal, randomized Phase III trial and increasing the likelihood of success. No longer can we justify inefficient early-phase clinical trials that are wasteful of the limited and precious resources now available to conduct clinical research in oncology.”
Patricia F. Dimond, Ph.D. ([email protected]), is a principal at BioInsight Consulting.
