It couldn't be a more confusing time for cancer patients. A year ago, results from early trials on humans raised hopes that a new class of drugs targeting tumor cells would bring the dread disease under control. But this year, continued testing has resulted in poor, or at best mixed, results for one promising drug after another. And the Food & Drug Administration has started telegraphing discontent with the trials themselves, saying it wants larger, more tightly monitored tests.
Just when it seemed as if this entire area of research would be seriously stalled, an FDA panel voted on Sept. 24 to recommend approval of AstraZeneca's (AZN) Iressa, a targeted therapy for patients with advanced lung cancer. The recommendation was made even though FDA staffers had expressed reservations about the data backing up the drug. The FDA almost always follows the advice of its panels.
The vote means that the many companies developing cancer drugs can breathe a bit easier -- but just a bit. "The take-away message from the panel vote is that targeted cancer drugs in the current environment are going to be few and far between," says Dr. Vincent Miller of Memorial Sloan-Kettering Cancer Center in New York, lead investigator of an Iressa clinical trial.
INCONCLUSIVE RESULTS. These targeted drugs, designed to attack cancer cells while leaving healthy cells intact, are proving devilishly hard to test. They can be extremely effective in patients with the specific cancer defect being aimed at, but identifying that subclass of patients isn't easy.
And no drug has demonstrated it can help people live longer, even if it's successful in shrinking the tumor. Combine these issues with the fact that companies are hoping to win FDA approval with very small trials -- that are sometimes sloppily administered -- and the stage is set for inconclusive results.
That's sure to be an ongoing issue for the 400 new cancer therapies now in clinical trials. Targeted drugs are designed to shut down one out of the hundreds of defective cellular mechanisms that drive uncontrolled cell growth, but most of these defects have no simple diagnostic tests.
SMALL SUCCESS RATES. Consider Genentech's (DNA) Herceptin. It can stop tumor growth for the 25% of breast cancer patients who carry a defective Her2 gene, which can be detected with a blood test. But "if that drug were tested on all breast cancer patients and not [only] those who expressed the Her2 gene, those studies would have been negative, too," points out Dr. Alan Sandler of Vanderbilt-Ingram Cancer Center in Nashville, Tenn.
Iressa's target is harder to track. The drug blocks epidermal growth factor (EGF), a protein that tumor cells feed on to grow and spread. Some 14 EGF drugs are in development, including leading candidates from Abgenix (ABGX), ImClone Systems (IMCL), and OSI Pharmaceuticals (OSIP). But none has worked in more than 20% of patients. Iressa significantly reduced tumors in only 10% of patients with advanced lung cancer in the trial data presented to the FDA, while another trial failed to show that the drug kept patients alive any longer than chemotherapy might have.
That may sound like failure -- unless you're one of the 10%. "My energy level is so much better. I play tennis now five to six times a week," says Marilyn Frank, 68, a Queens (N.Y.) patient who has been taking the pill daily for six months.
"DIFFICULT TO EVALUATE." FDA staffers complain that a low response rate isn't the biggest problem -- rather it's the sloppiness of the data backing up the claims about Iressa. In a report issued the day before the panel voted, regulators said that while there were hints of drug activity, the trial's design "makes it difficult to evaluate these results."
The report also complained that patients were included in the trial who didn't meet the proper criteria -- a charge famously leveled at ImClone's EGF drug Erbitux when the FDA rejected its application last December. In a prepared statement, AstraZeneca called the panel's decision "a vote of confidence in Iressa and the clinical benefit we have seen with this drug."
Sloan-Kettering's Miller says it isn't unusual for patients to slip into a trial who don't meet all the criteria. In Iressa's case, patients had to have already failed two rounds of chemotherapy. "It's tough to monitor dozens of different doctors in different sites around the world," says Miller.
"VERY CLEAR MESSAGE." It's also tough to evaluate a drug's worth without comparing it directly to a control group that received only a placebo. But most companies try to win cancer drug approval with trials of fewer than 100 patients and no control group -- results the FDA will accept, under certain circumstances. Still, Abgenix Chief Medical Officer Dr. Gisela Schwab says the FDA is "sending a very clear message that whenever we can, we should be doing large, randomized studies."
Such expanded trials may give a better picture of how well a drug works, but they'll also add considerable time and expense to the development process. OSI has already announced that it would double the size of its key clinical trial for Tarceva, an EGF drug it's developing with Genentech, from 330 to 700 patients.
Larger clinical trials won't solve all the problems, though. Because most tumors are caused by more than one cellular defect, no one targeted drug is likely to shut the defects down. Oncologists believe the best way to fight cancer is through combinations of therapies, but the FDA prefers to approve drugs that first prove efficacy on their own.
UNFAIRLY BLAMED? That may not make sense for drugs that are designed to work in conjunction with chemotherapy, as many targeted therapies are. "The FDA has to get more realistic," says Dr. Eric P. Winder, director of the breast-cancer unit at Dana-Farber Cancer Institute in Boston. "Most [drugs] are not designed to be home runs, and they shouldn't be judged as though they are."
The FDA complains that it's being unfairly blamed. "It's not the FDA's fault that there aren't more -- and better -- cancer drugs," says a top agency official. "There just aren't that many promising new drug candidates submitted for approval." Whatever the reason, patients are the losers. By Catherine Arnst in New York, with John Carey in Washington