Dendreon’s Provenge Works Best for Patients With Low PSA, Scientists Say
From the moment Dendreon started in business 20 years ago, most scientists have said that if its immune-booster for prostate cancer was going to work, it would probably work best at an early stage of disease, before tumors had gotten too powerful for the immune system to contain. Today, Dendreon is offering the latest slice of data that suggests that may be the case.
The Seattle-based biotech company (NASDAQ: DNDN) is announcing today a new analysis of its pivotal 512-patient study of sipuleucel-T (Provenge), which separated patients into four quartiles based on how high or low their prostate-specific antigen (PSA) scores were. The study wasn’t designed to get a statistically significant answer to this question, meaning the finding could be a fluke. Still, the results do align with what many scientists believe based on the history of clinical trials with the product—that it should work better among patients with less-aggressive forms of disease. The analysis, by lead author Gerald Chodak at Weiss Memorial Hospital in Chicago and colleagues, was posted online today on the American Society of Clinical Oncology’s website, in advance of its annual meeting.
“All of these analyses support the robustness of clinical benefit for patients, and this PSA quartile data is helpful for patients when you think about sequencing of therapy,” says Mark Frohlich, Dendreon’s chief medical officer. “The data strongly argues that using it as early as possible is the best for the patient and still allows you to go on and get other therapies.”
Dendreon won FDA approval back in April 2010 for the first-of-its-kind immune-boosting therapy, after the 512-patient study showed patients lived a median time of about four months longer on the drug than on a placebo, with minimal side effects. Since then, the company has faced a series of new competitive threats, particularly from Johnson & Johnson’s abiraterone (Zytiga) and Medivation’s enzalutamide. Those drugs are starting out by aiming to treat the sickest of prostate cancer patients, whose disease has worsened after getting chemotherapy. Dendreon’s drug is approved for patients with a less-severe form of disease, which hasn’t yet prompted them to go all the way to chemotherapy. Since prostate cancer is a slow-growing malignancy, part of Dendreon’s challenge is to persuade doctors to get more aggressive in treating patients early.
PSA, which isn’t a perfectly reliable marker, is still used almost universally by doctors and patients to track a patient’s progress. The lower the score, generally speaking, the better. In this analysis of the Dendreon study, researchers crunched the data to look at survival times of patients who entered the trial with PSA scores of less than or equal to 22.1; between 22.1 and 50.1; between 50.1 and 134; and over 134. As you can see from the chart below, Provenge patients lived longer than placebo patients in all four quartiles, but the difference in median survival times was the largest among those with lower PSA scores.
|PSA of 22.1 or less||PSA of 22.1-50.1||PSA of 50.1-134||PSA of 134 and up|
|Provenge||41.3 months||27.1 months||20.4 months||18.4 months|
|Placebo||28.3 months||20.1 months||15 months||15.5 months|
|Difference||13 months||7.1 months||5.4 months||2.8 months|
|–Source, Gerald Chodak et al|
This isn’t really a surprising finding, given that Dendreon has previously said that patients with lower PSA scores had better outcomes on Provenge, about to a summary of the trial published in the New England Journal of Medicine. But it’s the first time Dendreon looked more closely at the PSA scores, by dividing patients into quartiles, Frohlich says.
Given how much debate there has always been about the data to support Dendreon’s prostate cancer drug, I’d love to hear readers thoughts on how meaningful the latest PSA analysis is. Do you think the consistency of the survival advantage in all PSA groups means something, or do you chalk this up to mere data dredging that does little more than stir up a new hypothesis? Let me know your thoughts in the comment section below.