No Devil in Details: Dendreon Data Stands Up to Scrutiny from Doctors, Investors
(Page 2 of 2)
the survival benefit couldn’t have been explained by Provenge patients being healthier when they entered the study.
—The Kaplan-Meier survival curves showed the effect of the drug started early in the study, and stood the test of time.
—The company looked retrospectively for any abnormalities in subpopulations who might have done really well and skewed the overall results in favor of the drug. That analysis showed that all subpopulations of people on Provenge, regardless of their tumor aggressiveness, locations, and PSA diagnostic scores showed a consistent survival benefit, Frohlich said.
—The study failed to show a statistically significant benefit on a secondary goal—whether Provenge could help slow the spread of prostate cancer tumors. This matched up with the prior trial, and shows that slowing tumors isn’t necessarily a reliable predictor of whether a drug can help patients live longer. The measurement is difficult to measure reliably, which is one reason why it wasn’t the main goal of the study, Gold said.
—A little more than half of Provenge patients (54.1 percent) developed mild to moderate chills, compared with 12.5 percent on the placebo. Fever was the next most-common side effect, with 29.3 percent of patients reporting that event, which typically went away in one to two days, which is consistent with prior Provenge studies, Frohlich said.
—The number of serious adverse events in the study was almost identical between the Provenge group (24 percent) and the placebo group (23.8 percent). There were more serious fevers and pulmonary embolisms in the Provenge group, and there were more cases of pneumonia and deep vein thrombosis in the placebo group, although the overall numbers were small for all those effects.
—Patients in the placebo group were allowed to “cross over” and get a frozen form of Provenge after their disease progressed, because even though it might muddy up the waters of the trial, it would be unethical to deny a dying patient of a drug that might work. This sort of clinical trial design, assuming that Provenge is effective, might make the placebo arm appear to be doing better than it really is, and reduce the advantage for the Provenge group. Now that the trial is completed, all patients are eligible to get regular Provenge, Frohlich says.
The company plans to take its time to scrub through the records from the trial carefully as it prepares to turn in an amended application to the FDA in the fourth quarter of the year, Gold says. The FDA will then have as long as six months to complete its review of the application, he said.
Lots of questions about the business strategy are still to come, and Gold deferred most of them to an analyst day the company plans to host this summer. In terms of finding a partner, Gold repeated that the company hopes to market this product by itself in the U.S., and find a partner to help with winning regulatory approval and marketing overseas. The company also sees the Provenge method as validating its technology, which can be applied across a platform of other cancers, namely breast, colon, bladder, and kidney, Gold said.
“Clearly the people who treat this disease are excited about the data,” said David Miller, CEO of Biotech Stock Research, a Seattle-based research firm that began recommending Dendreon stock in 2001. “There are some people out there who still want to pick on it.”
News reports are pouring out of Chicago as some more skeptical doctors question how meaningful a median survival advantage of 4.1 months really is. Some might eventually raise a fair point of whether the drug is going to be cost-effective, given that it will undoubtedly cost in the tens of thousands of dollars per patient. Questions like that, and whether Dendreon can manufacture the drug efficiently enough to meet demand, hire the right people to market it cleverly, and win over some of the skeptics in oncology, are sure to be on Dendreon’s mind for months to come.
Trending on Xconomy
By posting a comment, you agree to our terms and conditions.