This is Part 1 of a four-part series. See also Part 2, Part 3, Part 4. View PDF of entire series.
In memory of Malcolm (Butch) Noonan, Part 4 of this series.
16 November 2010. The setting looked more like a United Nations venue than your typical meeting room or conference center—speakers arranged in an inner circle, each seat equipped with individual monitors and global power adapters, proceedings monitored discreetly from behind darkened glass in the back. But international politics was not on the agenda on 8 November 2010 at the European Medicines Agency’s (EMA) sparkling modern offices in London’s Canary Wharf business district. Instead, U.S. and European scientists, pharma researchers, and patient representatives spent a day with EMA regulators; officials from the agency’s transatlantic counterpart, the Food and Drug Administration, joined by video link at 6:45 a.m. Washington time, though some had been listening since 4:00 in the morning. The charge of the day was to begin a dialogue toward the shared goal of offering therapeutic treatment and prevention trials to families with autosomal-dominant Alzheimer’s (for extensive background on this form of AD, see Alzforum’s Early-Onset Familial AD section). These families are excluded from participating in such trials, even though they have contributed immeasurably to the progress of Alzheimer’s disease research over the past two decades and may harbor unique scientific potential for finding better drugs for all Alzheimer’s disease.
A treatment trial designed specially for this population would be unprecedented. “The really new topic here at the EMA today is treatment of the pre-symptomatic state. This is a prospect that has never been touched in the regulatory framework. It is a very new discussion, and today is a starting point,” said Cristina Sampaio. Besides being a clinical pharmacologist at the University of Lisbon, Sampaio serves on EMA’s Scientific Advice Working Party, which was established by the Committee for Medicinal Products for Human Use. Sampaio co-chaired the meeting together with Bruno Flamion and Kerstin Westermark of the EMA. Russell Katz, who directs the FDA’s Division of Neurology Products, led the FDA delegation.
This gathering was timely. It came not only on the day Eli Lilly and Company announced its purchase of Avid, a biotech company known for an amyloid imaging agent that is expected to receive FDA approval soon and to enable earlier-stage trials in Alzheimer’s disease, but also the day before the New York Times featured an article by its reporter Pam Belluck on the promise of such trials in familial AD in a special “What’s Next” Science Times section featuring “hot” fields in science. (For more information on the projects mentioned in the Times article, see ARF API series, ARF DIAN series, ARF ADNI update, and ARF ADNI series).
The EMA had organized the meeting in response to outreach by scientists of the Dominantly Inherited Alzheimer Network (DIAN). This collaboration of U.S.-British-Australian scientists to date has enrolled 125 of a planned 400 members of several dozen families with autosomal-dominant AD into a comprehensive six-year biomarker study of preclinical AD. From the get-go, DIAN scientists had intended to offer their study participants a treatment trial, and the London meeting represents the latest step in their effort. A prior meeting with industry representatives held on 23 March 2010 in Geneva, Switzerland, had made clear that the biopharma industry would consider making their drugs available for this novel kind of trial, but wanted to hear from regulatory agencies whether the regulators would support such trials and what kinds of trial design they would want to see. In Geneva that same week, DIAN scientists led by Randall Bateman of Washington University, St. Louis, Missouri, and EMA’s Sampaio, began planning a first such discussion. Initially expected to be small, the gathering grew when several dozen industry scientists requested a seat in the room. Maria Isaac of the EMA, who led the planning, told ARF that this level of industry interest is unusual, and had prompted the agency to host a larger event than originally envisioned.
Moreover, Huntington’s disease was added to the agenda because this neurogenetic degenerative condition poses similar challenges to autosomal-dominant AD. Having launched large observational studies as well as a systematic preclinical drug development effort, the HD field may soon be poised for similar kinds of trials in what Huntington’s researchers call “pre-manifest” disease. All told, some 70 people rushed through London’s pouring rain on a cold Monday morning, surrendered their dripping umbrellas at the door for instant wrapping in the agency’s smart “umbrella manager” machines, and settled in for a long day of discussion.
The upshot? Both the EMA and the FDA expressed great interest in such trials and took the same view on most of the big questions. Regulators rarely tip their hand in these exploratory conversations. Their guarded language can invite parsing and divining—a bit like the famously cryptic statements by the former chairman of the Federal Reserve, Alan Greenspan. Teasing aside, the FDA’s Russell Katz actually offered crisp guidance on one key question on which trialists need clarity, that is, which outcome measures the FDA would consider appropriate for a trial in pre-symptomatic participants. Because these people function normally, traditional clinical tools of showing whether the drug did the patient any good are useless. Katz said that a biomarker that has a lot of congruent evidence, even if it is not yet formally qualified as a surrogate, combined with a measure of subtle cognitive deficits, might be acceptable for such trials in ADAD. It would be preferable to have previously shown, in trials of the treatment in patients with sporadic AD, that there was a good correlation between the biomarker and a more traditional measure of cognitive function. However, even without that, the FDA would consider for review a trial that showed improvement in a relevant biomarker and in a sensitive measure of cognitive functioning in patients with inherited AD who did not have symptoms, according to Katz. Researchers took notice. “Biomarker plus a cognitive outcome—that, to me, was the clearest progress of the day,” said Stephen Salloway of Brown University Butler Hospital in Providence, Rhode Island, a DIAN investigator and member of its clinical trial committee.
In another positive note for families urgently waiting for new drugs, the regulators said that, given the limited number of study participants available for ADAD, one successful trial might suffice for approval. Typically, regulatory authorities require two such trials. But on a separate hope DIAN investigators are nursing—that the regulatory agencies would invoke an existing pathway called Accelerated Approval—Katz said the agency was not ready because it still had too few data to judge how strongly these biomarkers predict a future clinical benefit. However, to scientists parsing the oracle, the door on this option seemed ajar. “If the data from observational studies and drug studies using biomarkers continue to converge, we think this may yet become an option,” noted Bateman. DIAN scientists would prefer this pathway, because it gets drugs to patients faster. It enables a drug to receive the agency’s conditional blessing based on positive biomarker results, and allows the sponsor to demonstrate that the drug is clinically useful during subsequent Phase 4 observation. If those data come, the drug receives a full stamp of approval; if they do not, the agency withdraws approval.
Scientists asked regulators whether ADAD could be designated an orphan disease, as is HD. This disease category affords financial advantages to drug developers, and ADAD is rare enough to qualify. However, discussion with the EMA and FDA made clear that in order to receive this status, a drug would have to be presented as being specific to the orphan condition, and what the DIAN investigators really want is to develop drugs in ADAD as a steppingstone toward preventing all AD—late-onset and sporadic. Salloway said that, to DIAN investigators, orphan designation makes less sense than an alternative option Katz laid out, namely that of developing and approving a drug for ADAD as a specific subset of all AD. Baltazar Gomez Mancilla from Novartis Biomedical Research Institute agreed, saying: “Orphan indication is a secondary aspect.” (For more on these issues, see Part 3 of this series.)
A sticky issue in designing preclinical trials is how to handle placebo controls, and it proved contentious at the EMA meeting, as well. Some volunteers resent undergoing a study’s procedures without the hoped-for benefits of a drug. In both ADAD and HD, some say they would consider learning their genetic status if they could join a trial, but not if they end up on placebo. The Alzheimer’s scientists in London unanimously said their trials would use placebo controls to ensure scientific rigor, and offer drug to all participants after the blinded phase is over. To these scientists, the burning questions at this stage lie more in the specifics of whether they can design a trial that does not require participants to learn their genetic status, yet that does not accidentally reveal people’s status to them by dint of a drug’s known side effects. Detailed design questions such as these are where the DIAN scientists will focus their efforts next.
In contrast, the Huntington’s scientists were split between some who favored placebo controls and others who vehemently opposed their use. “Patients will never go for it; physicians will never go for it,” said Sir Michael Rawlins, who chairs England’s National Institute of Health & Clinical Excellence (NICE). Rawlins recommended the use of historical controls, whose past success he cited for the development of such drugs as penicillin, sulfonamides, and others. Likewise, Nancy Wexler, Columbia University, New York, who co-discovered the mutation that causes HD and has promoted research and awareness for the disease throughout her career, cautioned that patients might not comply with randomization. Both EMA and FDA representatives insisted on placebo, however. Without this built-in control, they said, trials would become uninterpretable and unable to support regulatory approval. “If you do a study, then it has to be one that is interpretable. If you do a study that you consider ethical but is not interpretable, that is not acceptable,” said Katz. Other HD scientists agreed with this stance.
Debate grew heated for a while but, in the end, generated a mutual understanding whereby a trial might forgo placebo controls for a drug that already had proven effective in manifest HD. In this scenario, trials in pre-manifest could conceivably proceed against historical controls. This is a more classical situation, and quite different from the new paradigm that had drawn the audience that day: that of testing unapproved candidate drugs in preclinical disease, Sampaio said. She noted that trials against historical controls only work when the drug’s effect size is large, and cautioned that neither HD nor AD research have to date produced a drug with a large effect size in clinically established disease. For the new paradigm of testing a drug preclinically that has not proven effective in clinical disease, placebo controls will be indispensible, the scientists and regulators finally agreed.
This discussion highlighted a difference between the HD and AD research fields. While the former operates from an assumption that drugs will help along the continuum of disease from preclinical to overt, AD researchers increasingly interpret a string of negative trials in mild to moderate AD to mean that drugs may well help at the preclinical stage but not once disease is apparent. “In AD, there is a concern that drugs given later do not work. There is a concern that promising therapies are failing in trials because they were too little too late. The disease appears to have a self-propagating, self-fueling, autocatalytic process, with inflammation and functional disconnection,” said Nick Fox of University College London. At least for anti-amyloid drugs, the best chance may lie in dousing the flames early. “It is increasingly clear that there is a very long pre-symptomatic period. What has moved us forward in our effort for trials now is the recognition from biomarker and imaging studies that we can see this period, characterize it, and stage it. It is our window of opportunity,” Fox said.—Gabrielle Strobel.
This is Part 1 of a four-part series. See also Part 2, Part 3, Part 4. View PDF of entire series.