Data from six additional months of treatment with an experimental drug for Alzheimer’s disease were enough to turn a failed mid-stage clinical trial into a possible success. On Wednesday, the Japanese pharmaceutical company Eisai and its U.S.-based partner, Biogen, explained how, providing details that point to ways that their drug, unlike many before it, might succeed in a larger test.
In a Phase 2 study, the Alzheimer’s drug BAN2401 reduced measures of amyloid protein buildup in the brain—this buildup is associated with the progression of the disease. The companies said that those results—across all doses of the drug—were statistically significant. Furthermore, the drug slowed cognitive decline by 30 percent in patients treated with the highest dose. These, however, were only secondary goals of the study (more about the caveats below).
Pierre Tariot, director of the Banner Alzheimer’s Institute, called the results “heartening,” but he cautioned that questions remain and more analysis is needed. Nonetheless, he said the data so far are encouraging for BAN2401, and for other antibody drugs in clinical testing for Alzheimer’s.
“This is a finding that is going to stimulate further investment about what anti-amyloid therapies will have what effects,” Tariot said. “I think it will stimulate more investment and more focused investigation.”
The results from the BAN2401 study, presented during the Alzheimer’s Association International Conference (AAIC) in Chicago this week, are being closely watched by investors, Alzheimer’s researchers, and patients alike. They represent one of the few times in clinical testing that an Alzheimer’s drug has shown that it might improve patients’ ability to function, not just clear up the clumps of proteins that are the hallmarks of the memory-robbing disease. Shares of Cambridge, MA-based Biogen (NASDAQ: BIIB) closed 2.95 percent higher than Tuesday’s closing price, then fell more than 11 percent in after-hours trading as the investment community processed the data.
The results come with important caveats. They are from a Phase 2 test that failed its main goal of improving patients’ cognition after 12 months. Eisai and Biogen continued the trial and found that the patients given the highest dose of BAN2401 saw their cognition—and other measures of their mental faculties—improve after 18 months of treatment, compared to placebo patients. However, Eisai and Biogen used two different methodologies to measure cognition in its study, which makes it more difficult to draw definitive conclusions about the drug.
The cognitive decline seen at the highest dose of BAN2401 exceeded investors’ expectations, wrote Leerink analyst Geoffrey Porges. But other details have caused confusion. The drug didn’t show a “dose response,” for example, meaning getting incrementally higher doses didn’t necessarily mean better results. The data “once again call the reliability of this result and any associated inferences into question,” Porges wrote.
BAN2401 is an antibody drug meant to break up the clumps, also called plaques, of amyloid protein that build up in the brains of Alzheimer’s patients. The popular “amyloid hypothesis” often used to explain the underlying cause of Alzheimer’s holds that breaking up these plaques and clearing them from the brain might slow, or even stop, the disease’s progression. BAN2401 is meant to bind to the clusters of proteins that go on to form amyloid plaques.
Yet every single drug meant to bust up amyloid plaques has failed in clinical testing, often after promising signals in early studies couldn’t be replicated in a final Phase 3 test. In the past two years alone, solanezumab from Eli Lilly (NYSE: LLY); verubecestat from Merck (NYSE: MRK); intepirdine from Axovant Sciences (NASDAQ: AXON); and Lundbeck’s idalopirdine have all failed late-stage studies. Those failures have challenged the amyloid hypothesis and prompted researchers to rethink whether targeting amyloid plaques is the right way to treat Alzheimer’s. But Eisai and Biogen’s report of success with BAN2401—and Biogen’s encouraging early results with another experimental drug, aducanumab— have, for now, kept hopes alive that an amyloid approach could work.
The Phase 2 study of BAN2401 split 856 patients—who had early Alzheimer’s and signs of amyloid—into six different groups. Five of those groups got different doses of the drug via a 60-minute infusion, and the sixth group got a placebo. The main goal of the study was to show a positive change, compared to placebo, in a score measuring their cognitive function. By that measure, the drug failed the study at the 12-month mark. But when Biogen and Eisai announced those results last December, the companies also said none of the conditions for stopping the study were met, so they decided to keep going to see if they could learn more.
Showing a meaningful benefit on cognitive function after 18 months of treatment was a secondary goal of the study. Other goals were positive changes, as seen in brain images, of amyloid levels in the brain and the volume of the hippocampus, the part of the brain associated with long-term memory. Earlier this month, Eisai and Biogen reported success on these fronts, without any significant safety issues. The most common side effects were considered mild: infusion-site reactions and brain swelling. Tariot said this swelling is expected in a treatment that works to remove plaque in the brain, and it’s seen in other antibody drugs, including aducanumab.
The BAN2401 study used an “adaptive” design, which allowed investigators to step in and tweak the study midstream, an approach intended to more quickly identify the best dose. Speaking at AAIC, Lynn Kramer, Eisai’s chief of clinical development and chief medical officer, said the top dose was given to the most number of patients. That dose was tested in 414 total patients, 253 of which received the drug monthly, and another 161 who received the drug bi-weekly. The drug’s effect on amyloid was strongest in the group that received this high dose bi-weekly, with 81 percent of these patients showing reduction in the protein buildup at 18 months.
One tweak made midstream was moving patients to different treatment groups. While these changes might have helped Eisai and Biogen more quickly find and test the best dose, they may have also affected the statistical analysis. Kramer said that in 2014, a health authority outside of the U.S. asked that patients who carry Apo-E4, a gene considered a risk factor in developing Alzheimer’s, not to be randomized to the bi-weekly, high-dose group. Also, patients who have that gene but had yet to reach six months of treatment were to discontinue treatment at that dose. Tariot noted that those steps might have removed patients most likely to show decline, affecting the results seen in this key test group.
The removal of the Apo-E4 patients prompted Evercore ISI analyst Umer Raffat to ask whether BAN2401 worked as well in Apo-E4 carriers as in other patients, and if it didn’t, whether removing those patients from the high-dose group might have actually pumped up the observed effect of the drug. Carriers of the gene did not respond as well to tests of earlier experimental drugs, he wrote in a research note.
Still, he noted some points that weigh in BAN2401’s favor and played down the role of the Apo-E4 gene in efficacy. If the gene was the key factor in the BAN2401 results, the patient group that received the highest dose of the drug monthly—which had more carriers of the gene compared to the placebo group—should have performed worse than those receiving the high dose biweekly. It didn’t.
Howard Fillit, chief science officer of the Alzheimer’s Drug Discovery Foundation, said he was not too surprised that the clinical trial had different results at 18 months than it did at 12. Alzheimer’s is a slowly progressive disease, and an additional six months of treatment might have made a difference, he said.
Fillit added that the study supports new approaches for enrolling and evaluating patients in clinical trials. He points to cancer, where patients are matched to clinical trials based on biological indicators. Fillit says that when he was involved in Alzheimer’s clinical trials in the early 1980s, researchers had no idea whether the patients enrolled in the study had the disease. Decades since, that approach hasn’t changed much. In Eli Lilly’s tests of solanezumab, brain imaging found that as many as 35 percent of patients in the studies had no amyloid plaque and did not have Alzheimer’s, he said. But in the BAN2401 study, measuring amyloid provided a biomarker useful for both enrolling patients and also evaluating the efficacy of the drug.
“A modern clinical trial [for Alzheimer’s] is going to be more like precision medicine, the way we do it in cancer,” Fillit said.
Though BAN2401’s results at 18 months are encouraging, its success was seen only on the secondary goals. Both Fillit and Tariot say the drug needs more testing. That means the companies will probably need to prepare for a larger Phase 3 clinical trial. Eisai’s Kramer said that the companies plan to talk with regulators to find out what will be necessary to move the drug forward.
Eisai and Biogen have been working together in an Alzheimer’s alliance since 2014. The companies are sharing in the development of four drugs, including BAN2401. Aducanumab, currently in Phase 3 testing, is another. And at AAIC, Eisai and Biogen also presented data from a mid-stage study of a third, elenbecestat.