A new analysis of biomarkers, superseding confusing results released in May, appears to have re-ignited enthusiasm for the experimental Alzheimer’s disease (AD) candidate, sumifilam, the lead development candidate at Texas-based Cassava Sciences Inc. Compared to a placebo, participants treated in the phase IIb trial with the filamin A modulator saw "significantly improved" biomarkers of AD as well as cognitive improvements, the company said. Shares (NASDAQ:SAVA) rose 133.4% to close at $7.75 on Sept. 14.
Monday's readout represented the final results of a post-hoc analysis of the study, a randomized, placebo-controlled trial that enrolled 64 participants. It replaces earlier results shared by the company that Cassava's president and CEO, Remi Barbier, said had indicated "huge dramatic swings" in biomarkers for AD that, across a month, appeared to follow a "drunken" walk." That initial reading suggested that the trial had missed its prespecified primary outcome, defined as a drug effect on cerebrospinal fluid (CSF) levels of tau protein and other biomarker assessments.
A second analysis showed a much different picture, including "robust statistical correlations in levels of biomarkers" over time, Barbier said. Asked later about whether culpability for the initial outcome had been assigned, Barbier demurred. "In the spirit of being gracious, we're not going to name names," he said.
The phase IIb study was conducted following a phase IIa trial that, while it showed reductions in neurodegeneration and neuroinflammation in AD, was critiqued as needing replication because it was an open-label study. In the trial, 64 people with mild to moderate AD, ages 50 to 85, were randomized (1-to-1-to-:1) to 100 mg or 50 mg oral sumifilam or matching placebo. Treatment was administered twice daily for 28 days.
Core markers of Alzheimer's pathology include elevated levels of tau (T-tau) and phosphorylated tau (P-tau 181), and low levels of amyloid beta42 (A-beta 42). Compared to placebo, patients assigned to receive a 50-mg dose of sumifilam saw a 15% decrease in total tau (T-tau) levels, and an 18% decrease in T-tau when given 100 mg of the study drug (p<0.01). Levels of P-tau also decreased relative to placebo, falling 8% lower in the 50-mg drug group and 11% lower in the 100-mg group. A-beta 42 levels increased, rising 17% for the 50-mg drug group and 14% in the 100-mg group.
In addition, Alzheimer’s patients treated with sumifilam showed directional improvements in validated tests of episodic memory and spatial working memory vs. patients on placebo, the company said. Cognitive improvements correlated most strongly with decreases in P-tau 181.
Sumifilam, a small molecule that has seen long-term support from the U.S. National Institute on Aging, has a dual mechanism of action that reduces both neurodegeneration and neuroinflammation by binding a single protein target, according to Cassava's senior vice president of neuroscience, Lindsay Burns. The drug's target is an altered conformation of the scaffolding protein filamin A, a large protein that interacts with more than 90 other proteins.
Burns, who discovered the target along with the company's scientific advisor, Hoau-Yan Wang, said that the altered shape of filamin A present in AD patients is critical to the toxicity of beta-amyloid. "It really acts as an accomplice to the toxicity of A-beta 42, the most toxic fragment of beta-amyloid," she said.
With the phase IIb study sorted, Cassava is continuing to conduct a long-term, open-label, multicenter, extension study of sumifilam, which is testing a 100-mg twice-daily dose for 12 months. The study's target enrollment is about 100 patients with mild to moderate AD, including patients from prior studies of sumifilam. So far, it's more than 50% enrolled.
"Ultimately, of course, our goal is to conduct a phase III study in Alzheimer's disease with sumifilam," said Barbier. Following an end-of-phase II meeting with the FDA, the company will develop a study plan for a phase III efficacy program and manufacture clinical trial supplies of the drug.
The late-stage program could start as soon as 2021, but "obviously these are very expensive studies," taking a minimum of two years, Barbier said. "We do not have the balance sheet right now to conduct [them]. Having said that, I do believe that with the strength and the magnitude and the consistency of our data, I don't believe raising the money will be an issue for us."
At June 30, the company reported having $25.3 million in cash and cash equivalents.