Prospects For Pfizer/Medivation's Dimebon Dashed By Phase III Flop

The drug's inability to show an efficacy benefit highlights risks of Alzheimer's disease drug development and pressures Pfizer to pull off some pipeline wins.

The Phase III failure of the high-profile Alzheimer's disease drug dimebon is a setback not only for Alzheimer's drug development, but also for Pfizer, which very much needs pipeline successes to offset upcoming patent losses, and for Medivation, whose near-term future hinges on the drug.

The companies announced the results of two highly anticipated Phase III trials of dimebon (latrepirdine) in Alzheimer's disease patients on March 3, and the results were dismal. In the CONNECTION trial, dimebon failed to meet either of two primary endpoints - measures of cognition and global function - or secondary efficacy endpoints compared to placebo. The other trial, a safety and tolerability study, confirmed dimebon's tolerability when used alone or with other Alzheimer's medicines.

The disappointing efficacy results came as a surprise because dimebon showed a strong efficacy signal in an earlier Phase II/III trial, published in The Lancet in July 2008. In that study, after a full year of treatment, dimebon-treated patients were significantly better than placebo-treated patients on key aspects of the disease. The primary endpoint, Alzheimer's disease Assessment Scale-cognitive subscale, was highly significant.

To some, however, the data seemed too good to be true. Dimebon, which has been approved since 1983 in Russia to treat allergies, works by blocking excess calcium flow through pores in the mitochondrial membrane, but how it works in Alzheimer's disease is unknown. The Phase II/III trial was small, enrolling just 183 patients, and conducted in just one country, Russia.

The setback highlights the high stakes game that is developing a drug for Alzheimer's disease. Though a risky prospect, dimebon could have offered a significant cash reward - it appeared to have a shot at being the first disease-modifying drug for Alzheimer's to hit the market. Moreover, it showed potential to improve disease symptoms, not just delay disease progression, which is the hope for compounds that are being developed to target amyloid beta.

Given the Phase III flop, the future prospects for dimebon now appear grim. Pfizer and Medivation have four ongoing Phase III studies evaluating the drug, CONCERT, CONTACT and CONSTELLATION, all studying dimebon in Alzheimer's, and HORIZON in Huntington's disease.

"With our colleagues at Pfizer, we clearly need to complete a comprehensive analysis of the CONNECTION data with a particular eye to determining what new information it offers that may impact our ongoing dimebon development program," Medivation CEO David Hung said in a same-day conference call.

Pfizer bought into dimebon in September 2008, paying $225 million in upfront cash plus milestones and assuming 60 percent of the development costs in exchange for a 60 percent share of the profits (Also see "Pfizer Finds Backup For Aricept In Medivation’s Dimebon" - Pink Sheet, 3 Sep, 2008.). For the first nine months of 2009, Pfizer paid Medivation $15.6 million in development expenses, according to a Nov. 4 filing with the Securities and Exchange Commission.

For Pfizer, A Pipeline Opportunity Shot Down

Dimebon offered Pfizer a chance to extend its Alzheimer's disease franchise beyond the acetylcholinesterase inhibitor Aricept (donepezil), which faces generic competition at the end of the year. And given the lucrative opportunity for a drug that can modify the disease, a dimebon success would have given Pfizer a pipeline win to offset the significant patent losses ahead, particularly for Lipitor (atorvastatin) in 2011.

Pfizer's next chance for success in Alzheimer's is bapineuzumab, a monoclonal antibody thought to induce an immune response to clear beta amyloid plaque buildup in the brain. Efficacy data on that drug, which Pfizer gained through the acquisition of Wyeth, has thus far been lackluster. The drug is now in Phase III and partnered with Johnson & Johnson after J&J bought rights to the program from Elan last year (Also see "J&J And Elan Strike A Deal On Key Alzheimer's Disease Assets" - Pink Sheet, 6 Jul, 2009.).

"Some compounds in Pfizer's pipeline appear much more 'real' than others do," Bernstein Research analyst Tim Anderson said in a same-day note. He had not included dimebon in financial forecast models. "With annual R&D investments of $10 billion, investors should expect - and demand - that a certain level of pipeline progress will be made."

Investors are especially wary after Pfizer lowered its 2012 revenue targets in February to $66-$68.5 billion from the $70 billion it forecast at the time it announced the Wyeth merger.

Medivation Investors Bail Out

For Medivation, the news is a significant blow. The stock closed March 3 down 67 percent to 13.10. Dimebon was Medivation's lead drug candidate, though not its only Phase III program. The company does have a drug, MDV3100, in Phase III trials for prostate cancer, which is partnered with Astellas. The company ended the third quarter with roughly $214.4 million in cash and short-term equivalents.

"We believe that in the near-term investors are likely to take a jaundiced view of the company because of the failure of dimebon, and we do not expect the negative impression of the company to change unless and until MDV3100 reports positive results in Phase III development," Hapoalim Securities analyst Raghuram Selvaraju said in a research note.

JP Morgan analyst Geoffrey Meacham noted that the dimebon failure in Alzheimer's diminishes the chance for success in Huntington's disease as well. "AD and Huntington's are very different diseases, but we think the fact that there was virtually no treatment effect in AD significantly decreases the probability that dimebon will be successful in Huntington's disease," he said.

Geography May Be One Explanation For Confounding Results

The CONNECTION study enrolled 598 patients with mild to moderate Alzheimer's disease at 63 sites in North America, Europe and South America. Patients were randomized to one of three treatment groups: dimebon 20 mg three times a day (TID), dimebon 5 mg TID or placebo TID for six months. No statistically significant improvements for the 20 mg TID group were seen relative to placebo on either of the co-primary endpoints, effect on cognition measured by the ADAS-cog or the effect of dimebon on independently-rated global function, measured by the Clinician's Interview-Based Impression of Change Plus Caregiver Input (Cibic-plus).

On the first endpoint, dimebon 20 mg-treated patients achieved a 0.1 point difference from patients on placebo. On the second, 64.5 percent of dimebon 20 mg-treated patients showed improvement or no change at week 26 compared to 65.4 percent of placebo-treated patients.

The big question that remains is why the results were so different from those seen in the earlier trial, published in the Lancet, especially since the two trials were so similar, including duration of treatment, patient inclusion and exclusion criteria and the primary efficacy endpoints, according to Medivation.

The key differences in design were the size of the trial and the different geographies in which they were conducted. The confounding nature of the findings was two fold: dimebon's performance in the CONNECTION study was worse than in the Lancet study, and the placebo response was higher.

Medivation's management said a subset analysis of the data is ongoing and that it will review geographic and trial site interactions.

Pierre Tariot, director of memory disorder at the Banner Alzheimer's Institute and an investigator in the trial, pointed out during the conference call, "The initial study was done in one country with a small group of sites, one language, and the second study was done in multiple countries in multiple languages. In a clinical condition that's so complex and where so many of the outcomes are based on language performance that certainly could be a concern."

-Jessica Merrill ([email protected])