A new set of guidelines released by the Collaboration for Alzheimer’s Prevention (CAP) aims to move the field toward greater sharing of data and samples from ongoing clinical trials, though the all-important details have yet to be ironed out. In the May 2016 issue of Alzheimer’s & Dementia, CAP members who help run four prevention trials with public-private partnerships, working with other stakeholders, lay out principles for sharing data and samples, and they encourage investigators of other trials to follow suit. For their part, academic and industry researchers alike support sharing study information, but voice concerns about ensuring the viability of trials.

“Traditionally in science, investigators keep everything to themselves. They protect what they are doing to gain a competitive advantage,” co-author Paul Aisen of the University of Southern California, San Diego, told Alzforum. “We are trying to move away from that model and toward the idea that science is most effective when it is entirely collaborative.”

As the Alzheimer’s field has shifted toward preventing the disease, many questions have arisen regarding when to treat, which biomarkers accurately assess disease progression, and how to adopt standard sampling protocols. If every trial sponsor worked alone, answers would come more slowly, design of new trials could suffer, and the field as a whole could be set back. Much can be gained by data sharing since, by necessity, prevention trials are lengthy, tracking biomarkers for several years. The hope is that by learning of both successes and failures, and analyzing larger data sets, other clinical trials can avoid repeating the expensive pitfalls of the past.

“There has been a strong sense that, without having that data and the samples directly accessible to the scientific community, a large amount of the learning that came out of clinical trials was not being translated to the rest of the field,” co-author and DIAN leader Randall Bateman of Washington University, St. Louis, told Alzforum.

CAP formed in 2011 as a voluntary initiative among AD and regulatory scientists to build consensus and harmonize their upcoming prevention trials so that they could advance the field at large (see Aug 2012 conference news). CAP now includes researchers from three trials co-funded by the National Institutes of Health (NIH), industry, and foundations: the Alzheimer's Prevention Initiative (API), the Anti-Amyloid Treatment in Asymptomatic Alzheimer's Disease (A4) study, and the Dominantly Inherited Alzheimer Network Trials Unit (DIAN-TU), as well as the industry-funded TOMMORROW study. Representatives from the U.S. Food and Drug Administration, the Alzheimer's Association, the National Institute on Aging, and the F-Prime Biomedical Research Initiative (FBRI) form part of CAP, as well. The current paper's first author, Stacie Weninger, directs FBRI and oversees Alzforum. The group outlined their goals in an earlier review (Reiman et al., 2016). 

The newly published guidelines suggest that researchers share screening and pre-randomization baseline data within one year of completing enrollment, and that post-randomization trial data be made available as soon as possible without compromising the trial’s integrity. The latter is particularly relevant for long-term continuation studies, as it may take several years to confirm that a drug that has been largely evaluated on biomarker changes during the initial treatment phase indeed confers a clinical benefit in people who were still cognitively normal when the prevention trial began. Additionally, CAP calls for all study data to be made accessible to the research community either after regulatory approval or within 18 months of the trial’s end, whichever comes first.

The paper also recommends that trials collect multiple biomarkers to help the field pinpoint the best ones, and that biological samples such as CSF and blood be made available to other researchers by request.

Participant consent forms for trials should also include provisions for the shared use of these samples. Co-author Reisa Sperling of Harvard Medical School in Boston, who co-leads the A4 trial, said that most people who participate in trials agree to future sharing of their samples and are happy to do so. “When you talk to most participants, this is what they want. They are volunteering for these studies, and they want the maximal benefit of their participation for the field,” Sperling told Alzforum. “That involves data sharing.”

Though researchers across the field are generally in favor of data sharing, concerns remain over how to maintain trial integrity. The authors encourage researchers to find new ways of enabling data sharing without, for example, accidentally unblinding treatment data or revealing the mutation status of participants.

In designing A4, Sperling said she insisted on creating provisions for data sharing. More than half of the participants have already enrolled, and Sperling hopes to be able to share baseline data within one year of enrollment completion. To make this possible while preserving the trial’s integrity, A4 triple-blinds the study participants. Subjects receive a code that is used by their study site, but when the data becomes publicly available, each subject will receive a new code that can’t be linked back to the original except by a handful of A4 researchers. While it still could be possible to identify some participants based on the size or shape of their skull on MRI or other imaging scans, scientists who access the data must agree to keep information confidential. Researchers have also discussed pooling data to help mask patient identity. “I think we are going to pretty extraordinary measures to make sure people are unidentifiable,” Sperling told Alzforum.

CAP is not the first initiative to promote data sharing. The Alzheimer’s Disease Neuroimaging Initiative (ADNI) also makes its data available to researchers, though being an observational study without investigational drug treatment means the stakes are lower. Sperling noted that her Harvard Aging Brain Study made baseline data available within six months after enrollment was complete. The Coalition Against Major Diseases, an initiative of the Critical Path Institute (C-Path) in Tucson, Arizona (see Dec 2010 conference news),  pools and analyzes clinical trial data to accelerate drug development. CAMD recently partnered with GAAIN, the data portal run by the Alzheimer’s Association.

The CAP investigators have not decided on a consistent format for sharing data. Each trial may take a different approach. Building databases and hiring statisticians who can process data requests will take time and money, and the source of funding for these efforts is as yet unclear. A4 and DIAN-TU have discussed using GAAIN; API has requested NIH funding to support sharing of data and biological samples using LONI, the Laboratory of Neuro Imaging at the University of Southern California in Los Angeles. The NCRAD, aka National Cell Repository for Alzheimer’s Disease, may house biological materials, and API has discussed setting up an independent review board to decide who should receive these precious samples. “Samples are a limited resource,” Andrew Satlin at Eisai in Woodcliff Lake, New Jersey, told Alzforum. “There needs to be an additional layer of thinking about sharing biological samples.”

CAP will need buy-in from the pharmaceutical industry for its guidelines to gain wide acceptance. “In general, the principle of sharing data in the field to help advance research into new treatments for Alzheimer’s disease is very much needed. And it is already happening to a number of public-private partnerships. It has been a great value to the field,” said Satlin.

While companies favor data sharing in principle, their actual cooperation depends in part on their confidence that data sharing gets the nod of the regulatory agencies, and that it will not hurt their trials, specifically Phase 3 registration studies. “In terms of clinical trials of investigational medicines, our first job is to maintain the integrity of the trial. If anything compromises the integrity of the trial, then it’s all for naught. That’s not good for the field, it’s not good for patients. It’s not good for any of us,” said Eric Siemers of Eli Lilly and Company in Indianapolis. Siemers said his company supports data sharing but is cautious about sharing information that might jeopardize its Phase 3 trials. He told Alzforum that Lilly would want reassurance from regulatory agencies in the United States, Japan, and Europe before making information public. He noted that data sharing in Phase 2 trials is more straightforward.

Billy Dunn, who directs the Division of Neurology Products at the FDA’s Center for Drug Evaluation and Research, is an author on the current paper and has spoken publicly in support of data sharing in general. Regulatory scientists have participated in CAP discussions since the group’s beginning days.

Satlin pointed out that while regulatory guidance is helpful, the responsibility to maintain trial integrity still rests with the researchers. “It’s up to the people who want to share the data to figure out a way to do it, and then to show the regulatory authorities that what they are doing doesn’t compromise the integrity of the trial,” he told Alzforum.

CAP organizers agree that while the specifics remain sensitive, overall public-private partnerships are moving toward the high bar set by the current principles. “Philosophically I think the industry is on board. There is more and more conviction among all stakeholders that a rising tide of sharing data will lift all boats,” said co-author and API co-leader Pierre Tariot of the Banner Alzheimer’s Institute in Phoenix. “A lot of work needs to be done in the next two years. It is one thing to articulate the principles, it is another thing to actually produce the data in a meaningful way. Let’s all pay attention to that,” Tariot said.—Patricia Waldron

Patricia Waldron is a science writer based in Dryden, New York.


  1. A perfect example for sharing biological samples is DIAN. This is exactly how biological samples from clinical cohorts should be made available to the research community with the goal of performing the best science. This is clearly in the interest of the participating patients and their families. The willingness to share such valuable materials may be exemplary for cohorts established in other countries including Germany.

    DIAN enabled us in Munich to obtain CSF samples from 127 well-characterized mutation carriers and 91 non-carrier siblings for an exciting study aimed at following microglial activation before disease onset and during disease progression. Sharing these samples not only allowed my lab to conduct a fascinating research project, but it also helped enormously to initiate true translational research in a group that had exclusively performed basic research for two decades.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Collaborative Umbrella CAPs Three Prevention Trial Initiatives
  2. DC: CAMD Convenes Stakeholders to Reform Alzheimer’s Trials

Paper Citations

  1. . CAP-advancing the evaluation of preclinical Alzheimer disease treatments. Nat Rev Neurol. 2016 Jan;12(1):56-61. Epub 2015 Sep 29 PubMed.

External Citations

  1. GAAIN
  2. NCRAD

Further Reading

No Available Further Reading

Primary Papers

  1. . Collaboration for Alzheimer's Prevention: Principles to guide data and sample sharing in preclinical Alzheimer's disease trials. Alzheimers Dement. 2016 May;12(5):631-2. PubMed.