Shares of Medivation Inc. and Dendreon Corp. took opposite trajectories Thursday after Janssen Research & Development LLC unblinded its Phase III study of Zytiga (abiraterone acetate) plus prednisone in asymptomatic or mildly symptomatic patients with metastatic castration-resistant prostate cancer (CRPC) who had not received chemotherapy.

The international, randomized, double-blind COU-AA-302 study conducted by Janssen R&D, part of Johnson & Johnson subsidiary Janssen Pharmaceutical Cos., compared 1,000 mg Zytiga once daily plus prednisone against placebo plus prednisone. The study enrolled 1,088 patients, with progression-free survival (PFS) and overall survival (OS) as co-primary endpoints.

Janssen R&D, of Raritan, N.J., unblinded the study based on the unanimous recommendation of the independent data monitoring committee following its planned interim analysis. The committee also recommended Janssen offer Zytiga to patients in the placebo arm of the AA-COU-302 trial, and the company plans to comply, according to Kellie McLaughlin, a spokeswoman for Janssen R&D.

The company plans to begin filing for regulatory approval in the U.S. and globally beginning in the second half of the year, McLaughlin told BioWorld Today.

Zytiga was widely expected to follow the path of its FDA-approved indication in metastatic CRPC following chemotherapy containing docetaxel. (See BioWorld Today, April 29, 2011.)

The Phase III study for the initial Zytiga indication also was unblinded at its interim analysis, based on statistically significant improvement in OS and an acceptable safety profile. A subsequent analysis with more mature data confirmed those findings. (See BioWorld Today, Sept. 13, 2010.)

At the time, analysts suggested that Seattle-based Dendreon, which markets prostate cancer vaccine Provenge (sipuleucel-T), still had time to cash in on its exclusivity. But that window may be starting to close.

Last fall, after Dendreon reported lackluster third-quarter Provenge sales, Credit Suisse analyst Lee Kalowski asked presciently, "Could this be as good as it gets? If Provenge is struggling to grow now, what happens if J&J's Zytiga posts positive pre-chemo data," perhaps as early as next year? (See BioWorld Today, Nov. 4, 2011.)

That may be the scenario Dendreon now faces, and those prospects were reflected in the company's shares (NASDAQ:DNDN), which tumbled 13.8 percent at the opening bell before regaining some ground to close at $10.11, down 76 cents.

But the short-term view is muddied by the fact that Zytiga's interim analysis did not observe statistical significance in OS.

"We had statistical significance for all PSF and all secondary endpoints, and we had a very strong trend for clinically meaningful survival benefit," McLaughlin said. As the OS data mature, the company will have more clarity, she added.

With placebo patients crossing over to Zytiga, however, crossover effect could potentially confuse the OS benefit even more.

That uncertainly offers a ray of hope – maybe more – for Dendreon, suggested David D. Miller, CEO of the independent research firm Biotech Stock Research LLC. Lack of statistical significance in OS in the '302 trial was "completely unexpected" in the wake of the strong OS data in Janssen's post-chemo COU-AA-301 study, he said.

Moreover, two survival analyses have been conducted in the'301 trial, and the hazard ratio worsened in follow-up analysis, Miller pointed out.

"If that trend holds true in the '302 trial – which could well happen because of crossover and other issues – the '302 trial will never be statistically significant and the drug will not be used in the pre-chemo setting except in very limited situations," he predicted.

Given the OS finding, Miller even questioned whether the FDA would approve Zytiga for chemo-naive CRPC patients, given the agency's penchant for focusing on survival benefit.

"And this is not some tiny little biotech that [the Oncologic Drugs Advisory Committee] likes to push around," he told BioWorld Today. "This is J&J."

Janssen plans to submit the total '302 data package to the FDA, McLaughlin said, but she declined to discuss the company's discussions with the agency over trial design or data interpretation.

On the whole, the Zytiga scenario is positive for Dendreon, Miller insisted, as did Reni Benjamin, senior biotechnology analyst at Rodman & Renshaw.

"Dendreon continues to target urologists, a strategy which, in our opinion, differentiates Provenge and ensures the longer-term growth," Benjamin wrote in a research update. "If Dendreon is successful in establishing Provenge as the standard of care among urologists, before the patient is handed off to the oncologist for all subsequent therapies, we believe Provenge sales could continue to grow and gain market share."

However, many analysts stridently disagreed that Zytiga's potential downside was good news for Provenge.

Dendreon "does not have an indefinite amount of time to make Provenge stick," J.P. Morgan analyst Cory Kasimov wrote in a research alert. "While we expected Zytiga to work in the pre-chemo CRPC setting, this result intensifies the pressure on DNDN to get docs hooked on using Provenge" before more traditional drugs hit the market.

And Jonathan Aschoff, Brean Murray Carret & Co. analyst, called Zytiga "the first of two shoes to drop on Provenge," predicting that "Zytiga's pre-chemo survival benefit will beat Provenge's benefit by several months."

No matter which side they fell on the Provenge debate, analysts generally agreed the Zytiga news bodes well for San Francisco-based Medivation and partner Astellas Pharma Inc., of Tokyo, which halted their Phase III AFFIRM trial of MDV3100 last fall in advanced prostate cancer following chemotherapy. (See BioWorld Today, Nov. 4, 2011.)

Medivation's Phase III PREVAIL trial, now under way, has enrolled 1 ,700 men with advanced prostate cancer who are chemo-naïve in a trial that is similar in design, but larger than Janssen's '302 study.

"We believe the positive '302 development represents a favorable read-through" for MDV3100's PREVAIL study, which involves a "highly similar trial design to '302 (same primary endpoints of PFS and OS) involving the same CRPC [patient] population," Jefferies & Co. Inc. analyst Biren Amin wrote in a research note.

MDV3100 and Zytiga share similarities in their general mechanisms of action, with MDV3100 theoretically having stronger activity through a broader scope of androgen-signaling inhibition, Amin added. MDV3100 also had a numerical advantage in the post-chemo setting, showing better OS (18.4 months compared to 14.8 months for Zytiga) and benefit over placebo control (4.8 vs. 3.9 months). Notably, MDV3100 does not require prednisone co-administration, "potentially translating to a real-world benefit," Amin wrote.

Investors took note, sending Medivation's shares (NASDAQ:MDVN) up 13.6 percent Thursday to close at $72.91.

The next inflection point in prostate cancer is likely to come at the American Society of Clinical Oncology (ASCO) annual meeting, scheduled for June in Chicago, where McLaughlin said Janssen plans to submit the Phase III Zytiga data.

"ASCO will be the key," Biotech Stock Research's Miller said. "I am certain [Janssen] will do another statistical run. If they present non-statistically significant survival data at ASCO, the drug is in real trouble. That's their coming-out party."