22 December 2011. In their quest to offer treatment trials to families with autosomal-dominant Alzheimer’s disease, the scientists driving the Dominantly Inherited Alzheimer Network (DIAN) have cleared the next two (or rather, one and a half) hurdles. They have corralled signatures on the dotted line from 10 pharma and biotech companies for a collaboration agreement that is unprecedented in the field of AD clinical trials. They have also submitted a grant application to the National Institute on Aging to support the first phase of treatment trials, together with private funds.

Until the agency reviews the proposal in early 2012, DIAN’s attempted leap over the funding hurdle will remain suspended in mid-air. On the pharma collaboration front, however, the push to do something truly new—that is, test experimental therapies as secondary prevention in people destined to develop Alzheimer’s dementia—has taken a definitive step forward. Before, the DIAN scientists and family representatives had convened various stakeholders to discuss the prospect of mounting such trials. At those meetings, pharma scientists had voiced intense interest, but also cited obstacles and remained non-committal. Since then, negotiations unfolded behind the scenes, and on 16 October 2011 in Washington, DC, a group of otherwise competing biopharmaceutical companies gathered after having formally agreed to help make DIAN trials a success. Each company has signed an agreement that it will jointly support the precompetitive aspects of DIAN clinical trials with time, expertise, and money.

“This is a milestone for DIAN,” said Randall Bateman of the Washington University School of Medicine, St. Louis, Missouri. “If you had asked me five years ago what the odds were that different pharma companies would sign the same agreement to get these prevention trials going, I would have said it would be hard to get one to do it. I would have said it is impossible to get 10 to do it. To me, that is a huge step.”

DIAN is a joint long-term project of families with autosomal-dominant AD and scientists in the U.S., Australia, and the U.K. It started in September 2008 with an NIA-funded observational biomarker study to characterize the natural history of AD starting in young adulthood. Yet from its conception, DIAN aimed to offer therapeutic trials to its participants. Since 2008, the momentum for secondary prevention trials in gene carriers, whose disease most closely represents the transgenic mouse models in which anti-amyloid drugs have been initially developed, has strengthened in the field at large in the U.S. and other countries as more researchers have come to view such trials as a unique opportunity to both prevent dementia and test the amyloid hypothesis.

While the pharma consortium was taking shape, DIAN has advanced on other fronts as well. On the national media front, it drew the eye of NBC’s Robert Bazell, who interviewed DIAN steering committee member Denise Heinrichs for the network’s Nightly News with Brian Williams last July (play video clip; for more on Heinrichs, see ARF related news story). On the science front, DIAN continued to enroll new families and measure their biomarkers and cognition, bringing the network up to 230 participants currently. Moreover, its scientists reached out to colleagues in Europe in an effort to make clinical trials available to additional families with autosomal-dominant AD there, whether they participate in DIAN itself or not. To that end, an online registry will enable clinicians and families worldwide to find out more about upcoming DIAN treatment trials starting in early 2012. And in preparation for a hoped-for start date of fall 2012, the scientists are designing a trial that would launch three drugs in parallel against a pooled placebo group prior to advancing the successful drugs into a larger trial. For details on the consortium, DIAN enrollment and data, families in other countries, and the trial design, read this update on all things DIAN.

The DIAN pharma consortium at present comprises Biogen Idec, Elan, Genentech, Janssen Alzheimer Immunotherapy, Lilly, Mithridion, Novartis, Pfizer, Roche, Sanofi-Aventis, with other companies potentially joining later. These are the companies that may have an unapproved drug they hope to see trialed in DIAN and were able to reach a common denominator with the DIAN’s academic home at Washington University, St. Louis, Missouri, on issues ranging from intellectual property to data sharing and funding. While the DIAN pharma consortium will assist with precompetitive issues, the DIAN Therapeutic Trials Unit (TTU) will develop and direct the planned trials.

How do the two groups relate? In essence, the consortium’s charge is to add pharma expertise to designing and implementing DIAN drug trials. The TTU controls and coordinates the trials with help from the federally funded Alzheimer’s Disease Cooperative Study (ADCS). Further, the TTU will manage a trial registry expanding into different countries and generally deal with patients and other groups such as regulators. The pharma consortium formed a biomarker working group that is analyzing DIAN data to help power trials adequately, as well as a separate group that works on identifying a cognitive outcome measure suitable to the presymptomatic and the earliest symptomatic phase of AD. Moreover, the pharma consortium funds needs, such as regular meetings with the TTU plus the expanded registry. It also funds a bank of fibroblast-derived stem cells currently being created from cheek swabs of DIAN participants for preclinical study of candidate drugs. Each consortium company may nominate a candidate treatment; however, the choice of what goes into DIAN trials is not theirs. DIAN’s TTU is evaluating the nominations and submits a short list to DIAN’s steering committee, who will make the final call.

What are the pharma companies getting in return? For one thing, a voice in shaping these trials, even if their drug is not chosen in the first round. A pharma company’s drug is no longer precompetitive, and, indeed, discussion filled an entire day without anyone naming a single treatment. The consortium agreement stipulates as much. But even so, each company involved struggles with similar biomarker issues as it develops its respective treatments. By participating in DIAN, they get to see data on how AD biomarker trajectories change reaching back 30 years before symptoms set in. This dataset complements what they are learning from the Alzheimer’s Disease Neuroimaging Initiative, which follows a different, i.e., non-genetic and older, population in otherwise directly comparable ways. Pharma company scientists can see natural history data on DIAN biological markers and cognition now, and they will see drug response data on the same markers as trials get going. For example, any company developing, say, an anti-Aβ antibody or a secretase inhibitor, would like to learn if, and how strongly, any drug in this class can budge CSF and imaging markers in presymptomatic or early-stage patients.

In preparation for trials, DIAN itself has continued enrolling. DIAN is an observational biomarker study that started in September 2008. By December 2011, 230 people had joined, of whom 64 have completed year 3, and 27 people year 4 follow-up visits. The DIAN scientists use the age at which a person’s parent became symptomatic to estimate when the adult child will develop AD. People aged from 15 years younger to 10 years older than their parent's age at onset are eligible to join a treatment trial. At present, about three-fourths of the participants are asymptomatic; of those, 85 percent are younger than their parent's age at onset. DIAN participants submit to what would for most people seem a grueling list of tests. But they are highly motivated. So far, 87 to 97 percent complete a long list of clinical, paper-and-pencil, and imaging assessments, and 76 percent consent to lumbar punctures.

What are those tests showing? At the October pharma consortium meeting in Washington, DC, Bateman summarized how mutation carriers compare to non-carriers. These are cross-sectional findings; longitudinal data are expected to come next spring. Clinically, mutation non-carriers are normal at all ages, whereas carriers subtly diverge from normal up to a decade before their parent’s age at onset. Roughly midway through this decade, their clinical findings reach a CDR of 0.5, and by their parent’s age at onset, carriers typically have a CDR of 2. Likewise, the MMSE starts trending down in carriers already a decade or so before. These emerging data look as though carriers might be getting sick somewhat younger than did their parents. In reality, the data may reflect the fact that carriers in DIAN are under such close observation that subtle deficits get picked up that in their parents were not recorded. The time point of five years earlier roughly corresponds to the stage at which incipient AD would be detected with a CDR of 0.5, said David Holtzman of WashU. These year estimates, and even the order of changes, are highly fluid at this point, as data are rolling in and being analyzed.

On the biomarker front, it appears at this early point in the study as if the CSF Aβ42 concentration is higher in young adult mutation carriers than in non-carriers of the same age (see also companion API story). At older ages, CSF Aβ42 declines in carriers. It transitions through the normal range starting at about -20 years, crossing lines with non-carriers (whose levels are similar at all ages) at about -15 years, and decreases further to levels typically measured in late-onset AD. “There is a strong trend to have elevated levels far earlier. Once we have a greater number of people and longitudinal change, we will know for sure,” said Anne Fagan of WashU. Fagan’s group measures all DIAN CSF samples to avoid the variation between groups that is seen worldwide and was recently confirmed in an ongoing quality control initiative (Mattsson et al., 2011). CSF tau was flat in non-carriers across ages but significantly higher in carriers by -10 years. As is also seen in people with sporadic AD, then, CSF tau in autosomal-dominant AD appears to rise some five years before a person becomes truly symptomatic, Holtzman said.

In plasma, Aβ is higher in carriers than in non-carriers; however, this measure appears to stay constant with age. To Fagan’s mind, this suggests that CSF reflects Aβ in the brain, whereas plasma reflects primarily Aβ from the periphery.

Brain amyloid as measured by PIB PET imaging crops up in the earliest affected regions, the precuneus and caudate, even at -27 and -25 years, respectively, Bateman reported. The only mutation carriers who are completely PIB-negative are the youngest participants. FDG PET measurement of brain activity in the precuneus trends steeply down at -12 years; unlike CSF Aβ42, FDG PET continues dropping as disease progresses. In toto, the biomarker findings available so far fit the story of autosomal-dominant AD being marked by Aβ overproduction, said Paul Aisen of the University of California, San Diego, a DIAN investigator and head of the ADCS.

Can More Families Join?
Autosomal-dominant AD is rare, and affected families tend to live scattered across many nations, not clustered conveniently near academic medical centers. Such families also speak of a sense of deep isolation. To help families connect, share experiences, learn about upcoming trials, and give their own input into trial planning, DIAN, together with the Alzheimer’s Association, has set up an online forum for families with this form of AD. Called the ADAD Forum, it opened in February 2011 and so far has 51 members. It facilitates commentary, offline communication, and periodic conference calls and Webinars.

On 20 November 2011, family members from Australia, the U.S., and the U.K. joined DIAN scientists in such a Webinar. They listened to the latest trial updates, asked questions, for example, about genetic testing and trial participation (testing is optional, not required), and gave researchers their feedback on certain trial design features. To protect ADAD families' privacy, this forum is not listed publicly, but people who believe this form of AD runs in their family and are interested in joining can contact DIAN study coordinator Wendy Sigurdson at sigurdsonw@neuro.wustl.edu for more information.

From the get-go, it was clear that the small number of participants would place restraints on scientists’ ability to power clinical trials. But as word about DIAN began to spread in other countries, scientists and families there became interested as well, opening the prospect of larger, more powerful trials. Last July at the 2011 Alzheimer’s Association International Conference (AAIC) in Paris, France, some 25 physician-researchers from a dozen European countries and Canada met with Bateman, DIAN’s principal investigator John Morris of Washington University, and other DIAN leaders for a dialogue on making DIAN trials more broadly available. As it turned out, 17 European research sites said at this gathering that they are working with 145 families including some 300 asymptomatic and 180 symptomatic individuals, many of whom may well be interested in participating in trials, Bateman told the pharma consortium at the October meeting in Washington. There may be even more families in Europe. In fact, at AAIC, researchers from several European groups independently presented data on autosomal-dominant families. For example, David Wallon from INSERM in Rouen, France, presented a poster with Dominique Campion and other geneticists from across France, which characterized 143 families with early-onset AD. Of those, 83 had a presenilin 1 mutation, 15 had an APP mutation, 12 had an APP duplication, and for 25 families the genetic reason for their disease is presently unknown. In these French families alone, 314 relatives are at risk for their parent’s disease, and 31 have chosen to find out their genotype after genetic counseling, Wallon told Alzforum.

It is not clear yet whether any additional European sites besides London will become bona fide DIAN sites, start up their own studies, or simply decide to offer DIAN treatment trial participation to their families regardless of whether that family participates in a requisite observation study. To enable trial participation, DIAN intends to launch an expanded registry website hosted at WashU with links from the DIAN website, from Alzforum, the Alzheimer’s Association, and national AD organizations in Europe. Through this registry, families with autosomal-dominant AD or their physicians can take first steps toward participating in a DIAN clinical trial. They can contact the DIAN research coordinator to obtain information; if needed, submit a saliva sample for confidential testing of whether an autosomal-dominant mutation indeed runs in the family; and then be matched to appropriate prevention and treatment trials within DIAN if they wish, Bateman said.

How to Test Drugs in This Unique Population?
Trial design took up much discussion at the pharma consortium meeting. In general, there was great interest in flexible designs that use run-in data from the observational period of DIAN to assess intra-individual change, as well as in trials with some adaptive elements roughly modeled after the I-SPY 2 process of breast cancer trials (see ARF related news story on adaptive trials). There are still many open questions and restraints. For example, it’s not decided yet up to what stage early symptomatic participants will be eligible. And even with the expanded registry, sample sizes will remain small enough so that most dose finding will have to have happened in company-run trials of each drug prior to the DIAN trials. That said, a draft design has taken shape, and the grant application DIAN scientists submitted to the NIA in October 2011 articulates it. The grant proposes a two-phase study backed by letters of support from the three companies whose drugs are named in the application, as well as support from the Alzheimer’s Association, ADCS, Alzheimer’s Disease Neuroimaging Initiative (ADNI), DIAN participants, and multiple external scientists. The first phase would determine whether the drug engages its intended target and whether it affects any downstream biomarkers of neurodegeneration. It would do that primarily by imaging whether amyloid deposition changes in response to the drug, and secondarily by comparing cognitive performance, biofluids, and other imaging markers before and after drug. The second phase would then look for a cognitive benefit of treatment, using primarily a to-be-determined composite panel of tests that are sensitive at that early stage, and secondarily, biochemical and imaging biomarkers of AD pathology and neurodegeneration, for example, MRI.

Importantly, the first phase would compare three different drugs to a shared placebo group. Each drug arm would enroll 80 people, assigning non-carriers to placebo to maintain genetic status blinding, and randomizing mutation carriers to drug versus placebo in a 3 to 1 ratio (a 75 percent chance of receiving drug). This design reconciles the twin dilemmas whereby participants may not want to find out their mutation status in order to join a trial, yet scientists do not want to randomize too many carriers to placebo or non-carriers to drug, either. That is because carriers lose time on placebo and non-carriers expose themselves to needless procedures, adding risk to themselves and cost to the study while adding little information to the study. In the proposed design, three groups of 30+ carriers each would receive one of three chosen investigational drugs, while the others would be pooled into one shared placebo group.

This first phase would go on for two years, at which point drugs that met primary aims would be considered for longer-term cognitive endpoint studies. Those drugs would then be tested in the entire population for three more years. Such a larger, longer trial is necessary for this second phase because its cognitive endpoints are likely to be subtle and change slowly in asymptomatic or very mildly symptomatic family members. If none of the three drug hits its target or a downstream biomarker in the first four-arm phase, then it would also likely fail to provide a cognitive benefit later on. Three new drugs would then be chosen for a second stab at Phase 1. In their grant, the DIAN scientists named three of the 12 drugs nominated to date by the participating pharma companies for this plan, but could not say which ones they are at this time.

The price tag for this plan? Sixty million dollars over five years. The NIA, so the DIAN scientists hope, will pitch in $3 million per year, and each sponsor of the three drugs would pay for one of the remaining thirds.—Gabrielle Strobel.

This is Part 1 of a three-part series. See also Part 2 and Part 3. Download a PDF of the entire series.