Some common themes emerge from the many failures of oncology drugs in phase III clinical trials.

On the road to approval by the U.S. Food and Drug Administration (FDA), cancer drugs often fail to make the near-final hurdle—meeting primary endpoints from a phase III clinical trial. In fact, only about half of the cancer drugs that undergo phase III testing go on to be submitted for FDA marketing approval, according to a 2010 review in Clinical Pharmacology & Therapeutics.

In most failed phase III trials, the new treatment doesn't improve overall survival. Such was the case with the recent POINTBREAK trial, which compared a 3-drug combination that included pemetrexed (Alimta; Lilly) to the standard 3-drug regimen for nonsquamous non–small cell lung cancer.

“We are disappointed with the results of this trial,” acknowledged Allen Melemed, MD, MBA, senior medical director with Lilly Oncology in Indianapolis, IN. “POINTBREAK did show an improvement in progression-free survival, though this did not translate to an overall survival advantage.”

Although results of each clinical trial are unique and open to multiple explanations, there are some common themes among the phase III stumbles for oncology drugs, says Kenneth Kaitin, PhD, director of the Tufts Center for the Study of Drug Development and professor at Tufts University School of Medicine in Boston, MA.

One theme is overly optimistic assumptions about a drug's potential when designing a randomized clinical trial (RCT). A 2012 Journal of the National Cancer Institute (JNCI)review of 253 phase III cancer RCTs published between 2005 and 2009 found that 62% failed to reach statistical significance.

Adaptive trial design, which allows investigators to modify the design of a study in progress, could help address this problem, the JNCI review authors assert. This strategy relies on rigorous statistical models to home in on early signs of success or failure and adapt accordingly—for example, by dropping a treatment arm, raising a dosage, or increasing the trial size. “Currently, adaptive trial design isn't widely used in phase III trials, but I expect we'll see more of it in the future,” says Kaitin.

Another problem among many trials is a lack of genetic analysis of individual patient tumors that can help to indicate which specific mutations may be driving the disease and thus which drugs may prove most useful.

Additionally, qualifying enough patients for phase III trials is difficult, and growing more difficult with the trend toward personalized treatments.

One promising solution that's becoming more common among trial recruiters, says Kaitin, is partnering with patient-centered organizations and advocacy groups to find appropriate study candidates. Pharmaceutical companies are also conducting more clinical trials overseas, especially in China, India, Eastern Europe, and South and Central America, where trial recruitment and retention rates tend to be higher than in the United States—due in part to economic, cultural, or other health care–related differences in those areas, Kaitin points out.

Overall, the most fundamental problem, he adds, is the sheer complexity of cancer.

“Cancer is rarely a single-mechanism disease, so improving survival with a single agent with a single mechanism of action is a huge challenge,” Kaitin says. Pharmaceutical companies are attempting to address this by testing multitarget inhibitors that attack cancerous cells by several mechanisms at once, he notes. However, it is early days for this strategy and it may come with a steep price in terms of added toxic side effects.