Shares of Dendreon Corp. popped up more than 50 percent in early trading on Thursday after the company clarified that either a positive interim analysis or a positive final analysis from its ongoing D9902B (IMPACT) Phase III study would satisfy the FDA's request for additional efficacy data on prostate cancer vaccine Provenge (sipuleucel T).

The interim analysis could occur "mid- to the second half of 2008," according to a presentation given by Dendreon President and CEO Mitchell Gold during the Bank of America 2007 Health Care Conference on Thursday.

While that's considerably better than waiting until the final analysis in 2010, it still sets the approval back by about two years. And it wasn't unexpected - even before the FDA's ruling, several analysts had predicted that the D9902B interim data would be needed for approval, and some voiced doubts that an approval was in the cards at all.

"There are people who just think this thing has a benefit, but I am not one of them," said Jonathan Aschoff, analyst with Brean Murray, Carret & Co. LLC.

During Thursday trading, Seattle-based Dendreon's shares (NASDAQ:DNDN) slid from their morning high but still closed up 27 percent, or $1.81, at $8.55.

With the most advanced therapeutic cancer vaccine in development, Dendreon has dragged its peers through the ups and downs of its recent regulatory roller coaster ride. On Thursday, South San Francisco-based Cell Genesys Inc. (NASDAQ:CEGE) closed up 25 cents at $4.33 and San Diego-based Favrille Inc. (NASDAQ:FVRL) closed up 6 cents at $3.97, while Redwood City, Calif.-based Genitope Corp. (NASDAQ:GTOP) closed down 8 cents at $3.57. (See BioWorld Today, April 2, 2007.)

Two months ago, when an FDA advisory committee gave Provenge the thumbs up, Dendreon skyrocketed from $5.22 per share to $12.93 and later rose as high as $25.25, with many of its peers basking in the reflected light. The Cellular, Tissue and Gene Therapies Advisory Committee voted 17-0 in favor of Provenge's safety and 13-4 in favor of its efficacy based on data from the D9901 and D9902A Phase III trials, neither of which reached its primary endpoint of time to progression. But a post hoc survival analysis of D9901 showed a 4.5-month survival benefit for Provenge-treated patients compared to those on placebo (p=0.012), as well as a 23 percent difference in three-year survival rates favoring Provenge over placebo (p=0.0046). (See BioWorld Today, March 30, 2007.)

Dissenting committee members urged the FDA not to approve Provenge, citing limitations of the post hoc analysis, the study's small size and confounding factors such as patient crossovers and chemotherapy use. Ultimately, the FDA delivered a complete response letter to Dendreon on May 8 requesting additional efficacy data and information regarding the chemistry, manufacturing and controls (CMC) section of the application. The stock plummeted to $6.33, pulling its peer group down with it while raising investor doubts about the regulatory pathway for active immunotherapy in cancer and setting off the usual slew of shareholder lawsuits. (See BioWorld Today, May 10, 2007.)

Now everything hinges on D9902B, a randomized, double-blind, placebo-controlled Phase III trial that has enrolled 425 out of a target 500 men with metastatic androgen-independent prostate cancer. Designed under a special protocol assessment with the FDA, the trial's primary endpoint is survival and its secondary endpoint is time to disease progression.

Gold said D9902B "is a well-designed trial and will definitively answer [the FDA's] questions." In his presentation, he noted that the interim analysis is well powered and will occur at an unspecified point after 164 deaths, the number observed in the integrated D9901 and D9902A datasets.

But analysts still are skeptical.

"I don't think this trial is going to work," Aschoff said. "It is far more properly designed and will be far more properly randomized than the only trial to show benefit so far." D9902B's 500-patient target is much larger than the 127 patients in D9901 or the 92 patients analyzed from the discontinued D9902A.

Stephen Dunn, analyst with Dawson James Securities, previously voiced concerns over the fact that past trials included only asymptomatic patients, while D9902B will include minimally symptomatic patients. When that issue came up at the Bank of America conference, Gold pointed to a Halabi Score analysis showing the patients in the D9901 and D9902B trials to be very similar in terms of predicted survival.

With its stock price on the ascent again, at least temporarily, Dendreon might take advantage of the opportunity to raise some money. The company filed a $200 million shelf registration in mid-March but hasn't yet offered shares, although Gold off-loaded some of his personal cache during the last price spike.

As of March 31, Dendreon reported $88.5 million in cash, equivalents and investments. The company projected its net cash to be used for operating and capital expenditures at $95 million for 2007 and $55 million for 2008. The lower costs next year reflect the completion of enrollment in D9902B and the fact that many commercialization-related expenses already have been incurred, such as inventory purchases and manufacturing facility costs. In May, Dendreon moved to further cut costs by laying off about 18 percent of its staff.