This is Part 1 of a three-part series. See also Part 2, Part 3. Read a PDF of the entire series.
14 February 2013. Over the past year, chatter has been building about combination drug therapy as the "New New Thing" in Alzheimer’s disease research. It is not idle talk. In the wake of anemic Phase 2 and 3 results, a movement has sprung up to turn ideas into action. Suggested by none other than Rusty Katz of the U.S. Food and Drug Administration, the topic of how the field could pull off clinical trials of two or more unapproved experimental drugs drew some 65 leading scientists and other stakeholders from across the country to Rockville, Maryland, last November. The occasion was the ACT-AD coalition’s fifth annual FDA/Alzheimer’s Disease Allies Meeting.
The scientists want to develop combination treatments with entirely new science. Rather than combine individually developed drugs in mild to moderate dementia, they are planning to test multiple experimental drugs in the preclinical stage of Alzheimer’s. “We are going to have to bite the bullet and be brave, and start combination trials as early as we can do it safely,” said Reisa Sperling of Brigham and Women’s Hospital in Boston, Massachusetts.
Dan Perry and Cynthia Bens of ACT-AD, the Washington, D.C.-based umbrella group for Alzheimer’s activist groups, hosted the meeting jointly with Diane Stephenson of the Critical Path Institute in Tucson, Arizona. C-Path is an applied research organization that engages regulatory with academic, industry, and other scientists to solve pre-competitive drug development problems.
After a day of discussion, the group parted with a pledge to take on the task. It is a large task, in part because testing drugs from more than one company in one trial or preclinical study will require a level of scientific and legal cooperation among companies that is nearly unprecedented in Alzheimer’s research. To facilitate such cooperation, the group agreed to whisk away leaders from the FDA, companies, universities, and other stakeholders for a three-day working meeting later this spring, where they would be charged with settling open questions and articulating a coordinated pathway that can be implemented. “Hermetically sealed room,” “Manhattan Project,” and “The Alamo” were some of the buzzwords that flew around, partly in jest, but partly in recognition of the size of the challenge.
Sperling urged that the spring meeting define specific action items that can be executed with current drugs and biomarkers. For his part, Katz said that the FDA wants to clarify with researchers exactly how a high-level draft guidance on combination trials that the agency published in 2010 can be applied to specific trial designs in Alzheimer’s. Michael Krams from Johnson and Johnson argued for a grander scheme, such as a jointly funded long-term biomarker observational study of up to 30,000 people. This study, “Framingham-on-steroids,” as Krams jokingly called it, would develop biomarker fingerprints for each disease stage going back to entirely asymptomatic, and then serve as a platform to spin off "sentinel" patient cohorts into a series of early-stage combination trials. Other speakers pointed to ongoing infrastructure in other indications that already enables multiple companies to jointly test combination therapies, or to contribute discontinued drugs to government repurposing initiatives (see Part 3 of this series). While people’s individual priorities varied, there was broad agreement that combination trials are both necessary and need a concerted effort to get off the ground. “We want to look back at today and say this is where it started,” said Stephenson.
Combination trials per se are nothing new. For years, researchers have tested two individually approved drugs together in one study, or added a single experimental drug to an approved one. For this approach, trial designs are established and the regulatory path is well trodden, said Owen Fields, a regulatory strategist at Pfizer. It also has yielded little improvement in Alzheimer’s treatment. For new AD drugs, this sequential approach of approving individually and then testing combinations is inadequate, scientists agreed. It is too slow, given that 10,000 baby boomers in the U.S. are turning 65 every day, said Stephen Salloway of Butler Hospital in Providence, Rhode Island. Other scientists expressed concern that some new drugs alone do not have robust effects on their own and would be discontinued even though they might well be effective when paired with another drug, especially in early disease stages.
Scientists are hoping to make a dent in the disease by combining two or even more disease-modifying therapies that are themselves in Phase 1 to Phase 3. This has never been done in Alzheimer’s research, though it is being done in cancer and being prepared in tuberculosis. And it can work. Last August, the FDA approved a combination treatment of four HIV drugs, two of which were previously approved but two that were new.
Several recent trends have converged to encourage researchers to try their hand at combination trials in Alzheimer’s. Setbacks in the clinic have reinforced the idea that it may take attacks on different pathways simultaneously to change the course of a disease as complex as Alzheimer’s. Research in tau and inflammation is rising in prominence and bringing these pathways to the fore as drug targets. Genetics research in AD is highlighting cholesterol management, endocytosis, and innate immunity as implicated in AD. “The risk genes tell us it’s not one gene, it’s not one target, and it’s going to take a multiple-target approach to be successful,” said Stephenson.
More narrowly, researchers who have conducted anti-amyloid trials say they have learned that hitting multiple targets even within just the amyloid pathway may be necessary to reduce amyloid levels earlier, safer, and more drastically than drugs have done to date. These researchers would start combination testing by adding a BACE inhibitor to an antibody. That is partly because for this type of combination, several Phase 2 or 3 drugs of both classes exist already.
Regulatory leaders have wanted to see combination trials ever since the FDA started the Critical Path initiative in 2004. Nine years later, both the will and the necessary tools appear to be in place. “We are very eager to develop combinations,” the FDA’s Bob Temple told the audience. Both Temple, a senior leader who oversees clinical science at the FDA, and Katz, who directs the agency’s Division of Neurology Products, spoke at this meeting (see Part 2 of this series).
The challenges are considerable. Starting with the scientific ones, Sperling noticed that it is unclear how trials can pick up combination drug effects, particularly in early disease stages, where the disease moves slowly. Biomarkers? They are proving their worth for selecting patients and indicating whether a drug hits its intended target. “But in terms of predicting clinical benefit, biomarkers are behind where we hoped they would be,” Sperling said. She was referring to the apparent dissociation between biomarker and clinical responses in the solanezumab and bapineuzumab Phase 3 results (see ARF related news story).
To Sperling’s mind, this problem can be solved by collecting biomarker information not only from small subsets of trial participants, but also from all participants in several more drug trials, such that scientists have more data to tell how each marker tracks with the subsequent clinical outcome. More sensitive cognitive measures are needed to detect the subtle changes that occur in early AD (see upcoming ARF Webinar on research in this area).
Getting Started with COMBAT
Gaps in those tools notwithstanding, Sperling has started planning a trial, to be called Combination Therapy in Early AD (COMBAT). The title is intentional. “After all, it is a war, and we are losing,” Sperling said. COMBAT could combine a secretase inhibitor and an immunotherapy to decrease Aβ production and facilitate clearance. In theory, such a trial could start in a few months, Sperling said. Alternatively, the trial could boost clearance by combining several antibodies that go after soluble, oligomeric, and fibrillar forms of Aβ. It could combine an anti-Aβ with an anti-tau drug to try to stem neurodegeneration, or add other neuroprotective or anti-inflammatory agents to any of the above. “I would love to build COMBAT to be a platform where we add COMBAT 2 and 3 as arms when these other drugs become available,” Sperling said.
The trial will contain adaptive features. Alas, which ones? Given the current lack of theragnostic power in the available biomarkers, it is unclear what short-term markers the trial could use to adapt dosing, randomization, or other parameters. One option would be to build early monitoring for ARIA-E into the model, because that side effect is known to happen soon after dosing, if it does. Another option might be to first run a small, continuous CSF monitoring study in some volunteers to determine when CSF biomarkers first respond to the given drugs—hours? days?—and build a lumbar puncture at that time point into the model as a basis on which to adapt. Importantly, the question of what markers to adapt on could be a deliberate objective of the trial, according to Don Berry, a leader in adaptive trial design at the MD Anderson Cancer Center in Houston, Texas. “You build the trial so it will answer that question along with other questions,” said Berry.—Gabrielle Strobel.
This is Part 1 of a three-part series. See also Part 2, Part 3. Read a PDF of the entire series.