or even a targeted antibody drug that’s supposed to seek out cancer cells and spare healthy ones. Instead, Provenge is designed to trigger the body’s natural immune defenses to recognize cancer cells as foreign invaders, like a virus, and kill them.
Dendreon’s approach requires blood to be drawn from a patient, and some white blood cells vital to the immune system, called dendritic cells, to be separated in a lab. The cells are shipped to the company and incubated with a genetically engineered protein found on prostate cancer cells, called PAP. This process is supposed to “teach” the immune system to recognize cells with this marker as foreign and fight them, and is sort of like waving a red flag in front of a bull. These newly revved-up white blood cells are shipped back in cold storage from the Dendreon factory to the clinic, and re-infused into the patient, giving them new ability to fight off the cancer. That’s the idea, anyway.
By 1999, Dendreon thought it had seen enough promising anti-tumor activity to take Provenge into the third and final stage of clinical trials, the kind of study that could pave the way for FDA approval to sell this product in the U.S. The gold standard measurement of effectiveness then, and now, for all cancer drugs would be to run a clinical trial that randomly assigns people to either the experimental drug or a placebo. Researchers would then follow patients for years to see if those on the drug live longer, with acceptable side effects in the eyes of the FDA.
Dendreon, like many biotech companies, couldn’t afford to wait around for the years it would take to show a survival benefit for patients with slow-growing tumors like those of the prostate. It might take years to enroll all the patients, and with their late-stage forms of prostate cancer, they still have a life expectancy of 18 to 20 months, so that would be a lot of waiting. So the company came up with a primary goal that would provide a quicker answer. They looked at how long patients on the drug are able to keep their tumors from spreading—a goal that’s long been thought to be a reasonably accurate predictor of improved survival time.
That first trial, of 127 men, failed to show the required statistical confidence that Provenge could slow down the spread of tumors, and for many biostatisticians, that is really where the story ends. It failed on its primary goal. Investors left the company for dead in 2002, driving the stock below $2. But Dendreon, again, like many other biotech companies in that kind of desperate situation, fished around in the data for something positive to hang onto. For a while, it tried to show a benefit for patients with slower-growing forms of tumors, without providing a benefit in the most aggressive cases.
That kept investors interested for a while, but Dendreon had another ace in the hole. From the beginning of that trial, the FDA required the company to keep track of whether these patients were alive or dead for a full three years.
It took time for this data to roll in, but the final analysis in 2005 showed that patients on Provenge ended up living a median time of 4.5 months longer than men in the control group, who got a placebo. About one-third of the men on Provenge were alive after three years of follow-up, about triple the three-year survival rate of men in the control group. The side effects were minimal—fever and chills that went away after a couple days.
Dendreon hypothesized that the vaccine just took a while to kick in, a so-called “delayed effect” that might explain why tumors would spread shortly after treatment, but the treatment would end up helping patients live longer after all.
Patient advocacy groups cheered on Dendreon from the sidelines, saying they needed alternatives to the existing treatments—chemotherapy and male hormone-deprivation—that have nasty side effects. Many scientists considered the data intriguing, but the sort of thing that would need to be verified in another, larger trial.
While all of this was going on, Dendreon had another trial ongoing called Impact, but that wasn’t really mature enough to produce a definite answer on survival. In consultation with the FDA, Dendreon agreed in November 2005 to expand the Impact trial to enroll 500 men. After all of these meetings with the FDA to talk about next steps, and before the results from Impact were ready, Dendreon decided to go for it. It completed an application to the FDA in November 2006, asking it to approve the drug primarily on the basis of the survival benefit shown in the earlier study of 127 men.
This gambit was controversial
