In 2011, as the Dominantly Inherited Alzheimer Network (DIAN), Alzheimer’s Prevention Initiative (API), and Anti-Amyloid Treatment in Asymptomatic Alzheimer's Disease (A4) trial platforms were each moving into place, their respective leaders realized they had much to gain from working together, and they formed an umbrella group called Collaboration for Alzheimer’s Prevention. CAP is a forum to coordinate development of new science needed for secondary prevention trials. Its scientists collaboratively try to ensure that its constituent trials gather data in a way that allows direct comparison, and seek public-private funding that facilitates sharing of data and samples with the field at large (see Aug 2012 news story). Having met regularly, this year CAP hosted the opening symposium at the 7th Clinical Trials on Alzheimer’s Research conference, held November 20 to 22 in Philadelphia, with an update on where their projects stand.
The Gordian knot of secondary prevention is to show a clinical benefit in people who have no symptoms. The many facets of the problem—from defining and recruiting trial populations to reducing measurement variability of biomarkers and creating outcome measures that will satisfy regulators—make it too big for an individual academic or industry group to solve. CAP reflects a realization that cutting the knot will require cooperation among many groups. “Secondary prevention does not take a village, it takes a federation,” quipped Pierre Tariot of the Banner Alzheimer’s center in Phoenix, the U.S. home of the API trials platform.
Besides leading clinicians of the DIAN, API, and A4 trials, the CAP umbrella includes Duke University’s Kathleen Welsh-Bohmer of the TOMMORROW study, as well as representatives of the National institute on Aging, the Food and Drug Administration, the Alzheimer’s Disease Cooperative Study, the Alzheimer’s Association, and the Fidelity Biosciences Research Initiative. (Fidelity provides financial support to Alzforum.) Each trial platform represents a large group of collaborators.
At CTAD, Randall Bateman of Washington University, St. Louis, cited as an example of CAP’s work its members' development of a cognitive outcome measure for their respective secondary prevention trials in close coordination with one another. New outcomes were needed because the CAP trials were enrolling asymptomatic or very mildly symptomatic patients in whom established cognitive batteries for mild to moderate Alzheimer’s disease were insensitive. This effort started with an API project led by Jessica Langbaum at the Banner Alzheimer’s center to analyze data from different longitudinal aging and AD cohorts in search of the handful of tests that changed the most during a person’s preclinical decade, but varied least from one person to the next. Soon after, Reisa Sperling of Harvard Medical School and Mike Donohue of the University of California, San Diego, started a similar project for A4, and Jason Hassenstab at Washington University St. Louis, Peter Snyder from Brown University, in Providence, Rhode Island, and others did one for DIAN. Each group, using different approaches in different datasets, arrived at largely the same composite of five cognitive tests that they are now using in their respective trials. At CTAD, Sperling said that each composite uses slightly different individual tests; the important thing is that each composite taps with equal weight into the same domains of cognition, such as episodic memory, attention, and executive function.
In CAP, scientists exchanged notes frequently on their respective efforts at devising new cognitive outcome measures for secondary prevention trials. [Image courtesy of R. Bateman, DIAN-TU.]
Moreover, the trial platforms jointly evaluate cognitive tests on iPads in hopes of gradually conducting more and more aspects of a prevention trial in people’s homes. Working with Dorene Rentz at Brigham and Women’s Hospital in Boston, the A4 group developed an iPad cognitive composite that contains some tests from the commercial CogState battery. They added a face-name task the Boston group has studied for years in preclinical populations using functional MRI, as well as a pattern-separation task that is sensitive to early hippocampal changes. The trials unit of DIAN (DIAN-TU) uses a similar iPad composite. DIAN-TU went a step further in transitioning parts of the trial into the home by hiring and training home health nurses near where trial participants live to visit and deliver study medication or placebo every month.
The ongoing DIAN, API, and A4 trials are all heavily loaded with biomarkers. CAP members have centralized data collection procedures for spinal fluid and imaging biomarkers. They standardized and shared protocols, as well as methods of analysis, Bateman said. “This is a huge deal to all of us,” said Tariot. “If we can show that a biomarker effect in two years powerfully relates to a clinical benefit in five years, then that opens the door for the whole field to conduct more efficient and shorter biomarker studies.”
All original CAP members have pursued public-private funding for their trials as a way to ensure that trial data, and even some samples, can be more readily shared across the scientific community than is typically the case with industry studies. National Institutes of Health and philanthropic grants frequently require that data and sample aliquots or scans be made available for other studies. The CAP members seek permission for this kind of sharing in the consent form they use when they enroll participants. They have worked out a process to release data in stages. For example, A4 screening data—including thousands of amyloid and tau PET scans—will become available when that study is fully enrolled, longitudinal data will become available when the study is complete, and treatment data when the regulatory process is complete, said Sperling. In the DIAN-TU trial, the sponsoring drug companies and the public-private DIAN-TU co-own the data and samples.
Some instruments, such as the cognitive composites, are already available. Another advantage of the public-private arrangement is that it can make a trial stretch toward perhaps-not-quite-achievable research goals, said Sperling. For example, A4 has mandated that 20 percent of screens be done in potential participants from an underrepresented minority. “This may be impossible, but an academic-private partnership allows us to try to understand what factors influence risk in African-American and Latino populations,” Sperling said.
Coordination notwithstanding, the trials differ in important aspects. For example, DIAN-TU, which enrolls people with autosomal-dominant APP or presenilin mutations across three different continents, has to the make the most of small participant numbers. It tries to maximize power from its current enrollment target of 210 people by using an adaptive design that tests gantenerumab and solanezumab against a pooled placebo group. The API’s Colombian trial, in 300 relatives of a kindred afflicted with the E280 presenilin (Paisa) mutation, is a single-center trial conducted in Medellin, Colombia, with a handful of satellite sites so people travel shorter distances for twice-monthly injections and safety checks. Both trials conceal mutation status, and so must include some non-carriers, all of whom are randomized to placebo. Both enroll easily, however, from an existing observational cohort of deeply phenotyped and motivated family members, hence almost no candidates fail their screen. DIAN-TU thus far has randomized 52 people at 20 sites, with one screen failure and no dropouts, Bateman said at CTAD.
In contrast, A4 aims to enroll 1,100 people with biomarker evidence of brain amyloid at 60 centers across North America and in Melbourne, Australia. The luxury of a much larger cohort allows this trial to be simpler in one way, said Paul Aisen of the University of California, San Diego. A4 is a straight-up parallel-group trial comparing one drug, solanezumab, to placebo. This avoids the potential statistical pitfalls that make some trialists nervous about adaptive designs until there is some experience with them in neurologic diseases. Rather, the challenge in A4 is to muscle through the sheer effort of screening. A4 needs to screen up to 10 people for each one it enrolls—yeoman’s work for both the sites and the potential participants.
Why so many screen failures? Sperling said a disproportionate number of people scored below the cognitive cutoff set in the protocol, largely for lack of appropriate adjustment to educational background. To address this, in one of several tweaks to the screening parameters, A4 broadened the range of cognition to allow more older people with a family history, an ApoE4 allele, and perhaps subjective memory concerns to become eligible for a subsequent screening PET scan, Sperling said. Other trials, too, are feeling their way toward the right cognitive and biomarker cutoffs as they screen pre-dementia cohorts.
At CTAD, Sperling said that to date A4 has screened 631 people. Thus far, 36 have received study medication and additional participants are eligible to be randomized. The first participant infused in June, in Providence, Rhode Island, allowed his infusion to be photographed by the Associated Press. “Some A4 participants have buried their parents with AD and have an older sibling in a nursing home. They are cognitively normal but are willing to go public and say, ‘I have amyloid and want to do something about it,’” said Sperling.
In 2013, Welsh-Bohmer joined the CAP group. She is the neuropsychology lead for the TOMMORROW trial of low-dose pioglitazone, which also enrolls participants at a pre-dementia stage of Alzheimer’s disease. Beyond that, this trial is different in many ways. Funded entirely by the companies Takeda and Zinfandel Pharmaceuticals, there is no public-dollar leverage to ensure early sharing of all data or samples.
Unlike DIAN, API, and A4—which ascertain amyloid positivity in their participants and have chosen investigational anti-amyloid drugs for their first trials—TOMMORROW is testing whether an approved diabetes drug can stave off a diagnosis of mild cognitive impairment due to AD. The study is based on the bioenergetics hypothesis of Alzheimer’s disease and posits that targeting mitochondrial, inflammatory, and metabolic abnormalities in Alzheimer’s pathogenesis with pioglitazone will delay progression. In addition, the trial aims to qualify a genetic biomarker risk algorithm developed from Allen Roses’ proposal of a TOMM40 length polymorphism (Roses et al., 2010). The trial splits cognitively normal people into a low-risk group and a high-risk group, treats half of the latter with 0.8 mg of pioglitazone daily, and then measures time to diagnosis of MCI due to AD in all three groups, Welsh-Bohmer said.
TOMMORROW is the largest trial in CAP. It aims to enroll 5,800 people between the ages of 65 and 83—4,600 for the high-risk group—at 60 sites across the United States, Europe, Russia, and Australia. This trial uses neuropsychology criteria for a diagnosis of MCI due to Alzheimer’s (Alberts et al., 2011) as its primary endpoint. Different sites in different countries implement MCI criteria differently, and the cognitive tests used in the trial have not been culturally validated; nor are preset normative data available for non-English countries such as Russia, Italy, Germany, or Switzerland. Therefore, the TOMMORROW team first conducted a validation and norming study. This pre-study has concluded enrollment, and sites in those countries will launch the actual trial in 2015, Welsh-Bohmer said. In English-speaking countries, the trial started in August 2013. Thus far, 37 sites have randomized some 1,000 patients with a screen failure rate of seven to one, she added. Welsh-Bohmer said that the TOMMORROW trial adds diversity to the CAP in its approach to risk stratification, the clinical endpoint selected for determining efficacy, and treatment.
Last but not least, while these four secondary prevention initiatives are out of the gate, a fifth is gearing up for a formal start in January 2015. At CTAD, Craig Ritchie of the University of Edinburgh, Scotland, U.K., told the audience that on November 14 the European Innovative Medicines Initiative (IMI) approved a handsome €64 million toward the European Prevention of Alzheimer’s Project (EPAD). EPAD is a public-private consortium to create a platform for a standing proof-of-concept trial of 1,500 people for the secondary prevention of AD. The idea is to gather some 24,000 people in a registry and enroll 6,000 of them into a longitudinal cohort of cognitively normal people who will be characterized with biomarkers and cognitive measures broadly similar to what API and DIAN-TU have been doing with their respective cohorts. Ritchie said that this biomarker phenotyping would be done to industry standards and create a reservoir population for intervention trials of people with biomarker evidence of AD who have very mild cognitive symptoms or none at all. The trials would evaluate proof of concept against a biomarker outcome of a succession of investigational drugs in multiple combinations of arms, emulating the principle of the I-SPY-2 trials platform for cancer.
Using an adaptive design, EPAD trials are to proceed in two steps. Any given drug is first evaluated against an intermediate biomarker that is linked to the mechanism of the drug—amyloid for an anti-amyloid drug, tau for an anti-tau drug. If it succeeds, the drug will go on to a second evaluation against a cognitive/clinical outcome. This outcome does not change from drug to drug and supports regulatory approval.
Even with €64 million, the project is so large that it cannot start from scratch, Ritchie said. He is reaching out to investigators of pre-existing aging cohort studies across Europe to invite them to collaborate. Ritchie called the amount of collaboration needed for EPAD to succeed “staggering.” He showed organizational charts and work package descriptions for how 35 partner organizations in academia and industry are going to work together. To pull the project off, EPAD will establish approximately 30 so-called trial delivery centers across Europe, starting with centers in Edinburgh, Scotland; Toulouse, France; Stockholm; and other cities. EPAD will formally kick off at a meeting on January 14 in Paris. —Gabrielle Strobel
- Roses AD, Lutz MW, Amrine-Madsen H, Saunders AM, Crenshaw DG, Sundseth SS, Huentelman MJ, Welsh-Bohmer KA, Reiman EM. A TOMM40 variable-length polymorphism predicts the age of late-onset Alzheimer's disease. Pharmacogenomics J. 2010 Oct;10(5):375-84. Epub 2009 Dec 22 PubMed.
- Albert MS, DeKosky ST, Dickson D, Dubois B, Feldman HH, Fox NC, Gamst A, Holtzman DM, Jagust WJ, Petersen RC, Snyder PJ, Carrillo MC, Thies B, Phelps CH. The diagnosis of mild cognitive impairment due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement. 2011 May;7(3):270-9. Epub 2011 Apr 21 PubMed.
No Available Further Reading