Where the Next Great Cancer Drug Might Be Coming From

"Phase 0" drug trials could cut R & D costs by hundreds of millions of dollars. So why isn't Big Pharma interested?


Where is the next great cancer drug? The number of new drug approvals has declined in the past few years, and despite rare breakthrough successes like Gleevec (for chronic myelogenous leukemia) and Herceptin (for breast cancer), most new "targeted therapy" drugs have offered painfully minimal gains in survival time. As a 2006 FDA report describes, in 1985, a novel agent entering the drug development cycle had a 14 percent chance of getting to market; by 2000, that chance had dropped to 8 percent. Meanwhile, the cost to take a drug from lab to patients increased by 55 percent from 1993 to 2003.

There's no sidestepping the painstaking lab work to find new ways to treat cancer. But there is a way to speed up development of potent new compounds and more quickly jettison the rejects: the phase 0 trial. As various technologies for watching what chemicals do once they get into our bodies have become available, researchers have gained the ability to test whether an agent they hope will target a specific cancer-generating mechanism actually hits its target. The cost of developing a new cancer drug is around a billion dollars. If a compound that hits its target in cultured cells and animals is shown not to do the same in a small group of humans, that could save hundreds of millions of dollars, not to mention several years. That is, if the pharmaceutical industry decides to play along.

The phase 0 study can turn the so-called "valley of death"—the chasm between studying a drug in animals and generating enough toxicology data and cash to study it in humans—into a few months' hike.

The process by which new anticancer drugs are tested is notoriously laborious. If a compound is active against cancer cells cultured in a lab, it is given to animals. If those tests look promising, human studies follow in three phases. Phase I "dose escalation" studies enable researchers to identify the maximum tolerated dose—that is, the highest dose at which the new agent can be given safely. Essentially, phase I ensures that a new drug won't kill a person in the process of making him or her well. In phase II, the drug is usually given in combination with an already approved treatment in a larger population of patients, usually at several centers across the country. Phase III are randomized trials involving a large group of patients and a "control" treatment, and are the last step before FDA review.

Although the phase 0 study has been around for a while, the concept has recently come into its own. The process was developed jointly by the National Cancer Institute and the Food and Drug Administration and was approved by the FDA in 2006. To be sure, there are many alternative trial designs (and funding for them), and collaborations between drug companies and the public sector, as well as academia, are on the rise. But the phase 0 study is especially promising. Because the goal is most commonly to confirm that the drug works as everyone thinks it will, a phase 0 study requires a very small number of patients—usually 10 or fewer—and very small doses of the drug. That's much more doable than even a phase I study.

The phase 0 study can turn the so-called "valley of death"—the chasm between studying a drug in animals and generating enough toxicology data and cash to study it in humans—into a few months' hike. Such was the case with ABT-888, a new drug from Abbott Laboratories. A phase 0 trial conducted by investigators at the National Cancer Institute confirmed that the agent, which was thought to inhibit the DNA-repairing enzyme PARP, was doing just that. The phase 0 study was ideal for testing ABT-888. The drug was designed to hit a certain cellular target, and the study, which enrolled 13 patients with advanced cancer, confirmed that the drug was going from point A to point B. As Shivaani Kummar, lead investigator of the study, explains, Abbott would have had to wait another year before launching a phase I study. After the successful phase 0, ABT-888 was able to go straight to phase II combination studies (several are ongoing).

The phase 0 study is not for all drugs. The approach is most appropriate "if there's a very straight yes or no answer," Kummar says. In other words, if confirming that a drug does what it was designed to do offers enough evidence to proceed to later phases—this isn't always the case, depending on the drug—then a phase 0 study is a logical first human trial. Phase 0 is also excellent for testing new delivery methods of existing drugs. Specifically, many intravenous drugs are now being converted to pills and other oral formulations, but swallowed compounds may affect the body differently from injected ones, a phenomenon easily evaluated in a phase 0 study.

The approach isn't perfect: if a drug fights disease in ways that researchers don't anticipate, a phase 0 trial can prove it ineffective even though it is actually beneficial. Nexavar, a recently approved drug, is a good example. It offers significant advantages in the treatment of liver and kidney cancers. Researchers initially believed that Nexavar worked by inhibiting an enzyme called B-raf. But the compound probably exerts its anticancer powers through other means. A phase 0 trial might have shelved this drug instead of getting it to the bedside of suffering patients.

And what's a pharmaceutical industry story without a little skullduggery? Drug companies, it turns out, may shun phase 0 trials precisely because they can be so effective. "People don't [always] want to find out if the drug is not working," Kummar says. It's not always advantageous to find out that your experimental agent doesn't work. A biotech company with a few drugs in the pipeline can keep investors interested with a "promising" new drug—and it's not that hard to show that a drug is promising—letting the company stay afloat for long enough to develop the next compound in its pipeline. Similarly, academic investigators in need of grant-earning publications would rather publish positive studies than negative ones.

But that's where the true value of the phase 0 trial lies: in halting, early on, the development of drugs that aren't any better than what's already available. The problem with the traditional phase I to III "critical path," as it's called, is that drugs often make it to phase II only to be proven ineffective, and sometimes even all the way to phase III. In fact, a new anticancer agent just failed a phase III study. "It turns out that it doesn't do what the company thought it did," says James Doroshow, of the National Cancer Institute, who declined to name the compound (the news will be public in a couple of months).

Still, the phase 0 trial probably won't bring down the commercial cost of drugs: the final price tag more closely reflects what the market will bear than the actual dollars spent on R & D. But the money saved could instead be directed elsewhere—toward, say, a dozen phase 0 studies instead of a single phase II. These trials would increase the odds of finding truly beneficial new compounds, potentially speed up the process, and spare a lot of patients dashed hopes. They might not lead to the next great cancer drug immediately, but they're a step in the right direction.

Image: Lyle Stafford/Reuters