By Gabrielle Strobel
Companies have not yet tried to repeat the statin story in Alzheimer disease. You cannot simply make a carbon copy of a successful drug development strategy, they say, and once you start thinking about the particulars of AD, tricky problems become all too apparent. "There is a sense that we should just do it and that is an understandable view, but when you think about the details it's an enormous challenge to do it well," said Dale Schenk of Elan Pharmaceuticals. That said, Schenk shares the sentiment of many investigators who consider the challenges surmountable. "There are biological issues, legal issues, ethical issues. They are complicated. And yet, the kernel of it that fascinates us all is, it would be such a great opportunity if we had a drug that is relatively safe and we could actually watch what happens in someone who we know has the disease process very early on," said Keith Johnson, who studies brain imaging in eFAD at Massachusetts General Hospital in Charlestown.
We describe below some of the important questions that scientists need to work out before they can design a good prevention trial. Almost every single one of these problems will come within closer reach of a solution as more people decide to volunteer for eFAD observational studies (see Essay 7. Larger studies mean more "power," or more statistically meaningful results.
Leon Thal, the late director of the federally funded Alzheimer Disease Cooperative Study (ADCS), estimated that a reasonably powered eFAD prevention trial would have to enroll 100 people. They have to have an identified autosomal-dominant mutation. Worldwide, between 500 and 1,000 such families are identified and known to various neurologists. (In the U.S. alone, more than 100,000 people are estimated to have early onset AD. But of those, many do not have a family history that indicates the presence of a mutation, or do not know their mutations, so no presymptomatic relatives can be identified.) Trial participants have to be of the right age, that is, within 10 years of the expected age of onset in their family. This way the trial would have a good approximation of when carriers would normally develop symptoms, and be able to measure a change in that rate. Not all of the participants in current observational studies at academic medical centers across the country and abroad would be eligible for a future prevention trial. Some already have AD and would instead need a treatment trial. Others are too young, and a trial will not be able to administer a drug for decades. Given those restrictions, there may not be many more than 100 patients accessible worldwide for a prevention trial within the next 5 years or so, Thal said. (By contrast, experimental proof-of-concept studies would not have to continue all the way through to expected time of AD diagnosis but simply measure a drug's short-term effect on biochemical and imaging markers. These kinds of study can be done with fewer participants of different ages.)
2. Who Runs It, Who Funds It: Public or Private?
A good prevention trial will require as much participation as possible from families who live dispersed all over the world. No single academic group, or company, is likely to be able to enroll enough patients on its own. Therefore, one option would be to organize a trial jointly between the ADCS (which works with 80 centers in the U.S. and Canada) and a similar European initiative called EADC, coordinated by French scientists. The Alzheimer Disease Neuroimaging Initiative (ADNI) is a large collaborative study that could potentially provide a framework for a joint effort on the observational phase that would precede a treatment phase, Thal said. To prepare for more international cooperation, U.S., European, and a nascent Japanese AD imaging initiative are already collaborating to ensure that their databases will be interoperable and able to support future drug studies, Thal added. Funding for such a joint initiative could come from the NIH and the European Union, or through creative industry/nonprofit partnerships, but these issues remain to be sorted out in the future.
3. Which Drug?
Different companies have different therapeutics at various stages of the clinical trials pipeline, plus some researchers are interested in certain approved drugs such as statins. Most AD trials fail, and scientists do not want to add to the burden of families with eFAD by putting them through multiple negative trials. Researchers from different institutions have suggested that a national registry be built, or at least some impartial entity be formed, that will help with these difficult decisions and guide families to the study best suited to them.
4. The Safety Bar: Is a Carrier "Healthy"?
A major reason holding back prevention trials is that none of the experimental drugs researchers are interested in has a sufficient safety record to be given to healthy people for a long time. However, familial AD challenges the definition of "healthy" and, by implication, of the safety standards that should reasonably be imposed on a candidate prevention drug. Arguably, a person whose presenilin mutation will make her demented within 5 years, and whose biomarkers are already changing, should be assigned a different health status than a random member of the general population, even though both people look outwardly the same. One has a time bomb ticking inside, the other most likely does not. AD researchers agree that AD develops through a long prodromal phase during which there are no overt symptoms. For a few years, a carrier is clinically "healthy" but pathophysiologically "ill." How safe does a drug have to be before it is ethical and appropriate to test it in an eFAD carrier? The FDA is charged with protecting people from hazardous drugs. Must it not also allow reasonable attempts to avert certain disease? What's the right formula for eFAD carriers? AD researchers say in unison that treating presymptomatic patients with drugs that have side effects crosses into new territory on which they need guidance from the FDA.
5. How to Measure Efficacy?
How would a drug developer know whether a drug is slowing down Alzheimer disease in a person who is at the earliest stages of illness? The disease moves slowly, symptoms are subtle, and they vary from person to person, from day to day. But techniques to measure changes are improving both in neuropsychometrics (tests of memory and other brain functions) and brain imaging and other biomarkers. Scientists would need to come to a consensus on how they will combine these measurements to detect a meaningful response to treatment at the earliest stages of disease. They also need data to satisfy the FDA that certain biomarker readouts are linked to clinical improvement.
6. How Long Do You Treat?
Will presymptomatic carriers take the trial drug for 6 months? One year? Two years? At the end of the study, do they stop taking the drug? Carriers are at 100 percent risk—is it ethical to deny them drug? Some trials offer open-label extensions so the company can gather long-term tolerability data. Would such an arrangement be feasible and ethical?
7. What Is Success?
A prevention trial aims to avoid future disease. How will we know a drug works? Each family with eFAD has an age range in which AD tends to become manifest. In some families this range is between 42 and 48, but in others it spans 20 years. Companies don't like to finance trials that take 25 years to complete. Economic realities demand faster drug development timelines, and such trials would instead lengthen the time to approval. Private or public sponsors will need to define by what criteria, short of a normal life span lived without an AD diagnosis, they will declare success or failure.
8. The Ethics of Knowing
Many people in eFAD families do not want to know their genetic status. Can a well-designed trial honor that wish? A double-blind randomized trial has the least bias and is considered the gold standard. If such a trial honors the wish not to know, the trial design would mean people who are carriers take a drug that has its own risk of side effects. But it also means that people who do not carry the mutation still expose themselves to the risks of taking the drug. One alternative is that people who wish to participate in the trial have to agree to know their status. Another is to include both carriers and non-carriers in research studies but withhold genetic knowledge from them, and to run a half-blinded drug study in which only carriers are treated with active drug while non-carriers receive a placebo. The downside of this approach is that people who develop drug side effects would become aware they carry the AD mutation by virtue of having the side effect. (This issue has in past trials been addressed by designing a placebo with side effects that are not dangerous.)
Presymptomatic carriers of eFAD mutations are vulnerable to insurance and employment discrimination. Even if individual carriers are open about their carrier status, a sibling, daughter, or nephew may desire more privacy about the family's genetic status. A trial must solve the legal and logistic problems of protecting people's privacy throughout the process.
10. Family Support, Enrollment
A prevention trial will need to supply the logistic and staff resources to locate people and support them through the decision-making process of whether they want to participate and whether they want to learn their genetic status. Some of these resources are available through the larger observational studies at academic medical centers and at some ADCs. This requires time and sensitivity, as family dynamics can complicate the communication. Study coordinators and neurologists often get instructions from a given person about whom they can and whom they cannot talk to about the person's genetic status. Families' communication barriers do not simply fall away because they face a common burden, and that form of within-family privacy must be respected, as well (see Genetic Testing and Counseling for Early Onset Familial Alzheimer Disease).
Taken together, pharma researchers say their companies will not now invest in prevention trials. Population-based prevention trials are too risky and costly, and eFAD prevention trials by themselves would not allow a company to register the drug for prevention in LOAD, the market they care about most. Companies will focus first on treatment in the large sporadic LOAD market, but they may be willing to donate study drug for a parallel prevention trial and support it in other ways. That could be very valuable, and scientists across academia, the NIH, and industry are professing interest in learning more about the challenges of such trials.
It seems clear that progress toward prevention and toward more opportunities for families with eFAD will require scientists to take collaboration to a new level. They need to share patients, harmonize observational studies so data can be compared side by side, and tackle trial design creatively. John Ringman, a physician-researcher at University of California, Los Angeles, reflected a commonly expressed sentiment by saying, "Competition is good; it's how we make progress. But we all realize this has to be a collaborative project. I am willing to take a step back if we can accomplish something unique."
Finally, researchers say that the FDA has to support prevention trials with guidelines tailored to the specific situation of eFAD carriers. Funders have to step up and work out arrangements of cost sharing. Lastly, patients and families who want to join the fight need to be given ways to do so. To date, individuals run the gamut from openness to denial. Some are "out" and waiting desperately for a drug trial, others guard their privacy but jet around the country volunteering for multiple studies, still others avoid the issue entirely. Many need support, counseling, and time to make their individual decisions. Importantly, genetic testing and trial participation are separate issues for them. Some at-risk relatives have said they can't face knowing their status until there is at least a trial for them. For others, it's the opposite. "For each subject in my study, I have contacted four people who know they carry the mutation but don't want to do research. They prefer to wait and see what happens to the other subjects," said Daniel Pollen, a physician-researcher at the University of Massachusetts Medical School in Worcester. "I hope that once the first studies show that a treatment at least reduces relevant biomarkers, their stance will tilt…to proactive presymptomatic treatment. I do think the climate is changing in favor of knowing and participating."
In the next, and final, installment of this series, we will investigate the options today for participating in observational studies of early onset familial Alzheimer disease.