When 715 scientists gathered in Philadelphia for the 7th Clinical Trials on Alzheimer’s Disease (CTAD) conference, their presentations and debates reflected a field that is revamping itself in hopes of more success in the next wave of drug trials. For the past decade, clinician-researchers have discussed innovation and laid the groundwork; however, in practice most intervention trials still hewed to the conventional formula of parallel group studies in clinically diagnosed mild to moderate Alzheimer’s, with little use of biomarkers. Now, all that is changing. “Things are completely different from five years ago in the drug development process,” said CTAD co-organizer Paul Aisen of the University of California, San Diego, who heads the Alzheimer’s Disease Cooperative Study (ADCS).
Even as change is unfolding, however, it is too timid for some. The best-received talk at CTAD was a keynote by Rusty Katz, who directed the FDA’s neurology division until last year. Katz exhorted the field to stop spinning its wheels trying to earn a label of disease modification. Rather, trialists should focus on delivering a large therapeutic effect, regardless of whether the drug changes progression or symptoms. That would honor the therapy goals of the National Alzheimer’s Plan, he said, and may require co-developing several investigational drugs together (see Part 2 of this series).
By now, researchers across the field fully recognize the importance of the roughly 15-year preclinical phase of Alzheimer’s. “DIAN [the Dominantly Inherited Alzheimer Network] and API [Alzheimer’s Prevention Initiative] have shown us that there is a sequence of biomarker appearance, and AIBL [the Australian Imaging, Biomarkers and Lifestyle study] has helped us have a lot of faith in this sequence,” Rachelle Doody of Baylor College of Medicine in Houston said in her keynote address. Prospective studies are largely convergent in their finding that brain amyloid predicts cognitive decline, albeit at rates that are subject to other risk factors, as well (see part 11 of this series.) Trialists have begun to harness this long preclinical window of time for therapy development. This means new ways of doing things across all steps of the process. Secondary prevention trials—in people who have evidence of AD biomarkers but no or very subtle symptoms—have begun enrolling based on modern diagnostic criteria, and their leaders are swapping notes on what needs tweaking as first experiences roll in. Leaders are banding together in groups such as the Collaboration for Alzheimer’s Prevention (CAP) to coordinate their searches for outcome measures and biomarker procedures for such trials. The DIAN, API, and Anti-Amyloid Treatment in Asymptomatic Alzheimer's Disease (A4) trial platforms have inspired the European Prevention of Alzheimer’s Project (EPAD), an international consortium to build a standing platform of secondary prevention trials (see Part 3 of this series). Registries are springing up in the United States and Europe to help these large studies recruit asymptomatic participants. “The idea of secondary prevention has become mainstream thinking,” said Aisen.
Even as the first secondary prevention trials are already enrolling, it is clear that some of the newer techniques they are using need refinement as the trials proceed. For example, a Phase 2 trial of Genentech’s therapeutic antibody crenezumab showed that amyloid PET may require a different reference region than used before—and than specified in current prevention trial protocols—if PET is to reliably gauge a drug effect in longitudinal measurements (see Part 4 and Part 5 of this series). Scientists know that using evolving methods in trials poses risks, but they have to do so anyway because following tried-and-true methodology has led them down dead ends for the past 20 years.
Despite being a truly nascent technique, tau PET is expected to be included in secondary prevention trials as soon as January 2015. “Tau PET is clearly a game changer,” Aisen said. Ask any trialist, and they express equal excitement about being able to both image defining Alzheimer pathologies in living people and to test more tau-based drugs (see tau therapeutics). As researchers scan observational cohorts and compare results against Braak staging of tau pathology, a unifying idea is rapidly gaining currency: that some tau pathology accumulates in most people’s medial temporal lobe as they age, but in the large subset of older people who accumulate amyloid, the tau pathology changes from a circumscribed process of aging to a propagating lesion that claims synapses and brings on AD symptoms (for more on tau, see Part 6 of this series.)
Researchers are stacking their initial secondary prevention trials with exploratory cognitive measures in hopes of moving beyond the paper-and-pencil tests that trial volunteers have been taking for decades. Home-based iPad tests and dementia staging tools, as well as patient-reported outcomes, already are being validated to generate richer datasets even while relieving study volunteers of the burden of having to trek to their memory clinic as often (see Part 7 of this story).
The next frontier, in preparation for primary prevention trials, will be mining large-scale, Web-based gaming data or unsupervised cognitive assessments taken by tens of thousands of people. This could help clinicians spot subtle decline brought on by soluble forms of amyloid. “Remember, we do not think fibrillar amyloid is the major driver in Alzheimer’s disease,” said Aisen. “If that is true, then we should be able to find a cognitive indicator of soluble amyloid toxicity prior to big masses of insoluble deposits spreading across brain and fueling big masses of tau spreading.”
That is the future, however. The present, on display at CTAD, features innovation in early stage trials—innovation that was guided in part by recent revisions of the diagnostic criteria for Alzheimer’s disease. Leading an international group of clinicians, Bruno Dubois of the Pitié-Salpétrière Hospital in Paris first introduced new criteria in 2007, then proposed a new vocabulary for what to call the early stages in 2010. “These attempts at nosology have been very helpful at stimulating a thorough discussion,” Doody said. This past summer, Dubois and the international working group further updated the criteria in light of longitudinal biomarker data gathered since 2010 (see Dubois et al., 2014).
At CTAD, Dubois said that the new criteria use a single framework for the range of preclinical and symptomatic AD. They require biomarkers as part of the diagnosis at each stage. The criteria split the biomarkers such that only the CSF signature and amyloid PET remain part of the diagnosis. In essence, the pathophysiological markers are integral to the Alzheimer’s diagnosis, but volumetric MRI and FDG PET got kicked out. Why? The former are specific to AD, the latter are not, hence they are less useful for diagnosis. Rather, volumetric MRI and FDG PET are suited for staging and monitoring progression because they stay dynamic as disease progresses, Dubois said.
The new Dubois diagnosis always has two components. The biomarker component requires either the pathological CSF signature of both Aβ/tau or amyloid PET (or an autosomal-dominant mutation), while the clinical component varies with the person’s presentation. For example, a diagnosis of typical AD requires at least an episodic memory impairment—isolated or with other cognitive or behavioral changes—that got gradually worse, plus the biomarker component. A diagnosis of atypical AD requires a clinical phenotype of either the posterior, logopenic, frontal, or Down’s syndrome variant of AD, plus the same biomarker component. A diagnosis of mixed AD requires an AD clinical phenotype with pathology biomarker to establish that AD contributes to the mixed disorder, plus clinical and biomarker evidence of either cerebrovascular or Lewy body disease.
For the diagnosis of preclinical AD, Dubois et al. maintain a distinction that came into question at CTAD. Among aging people with biomarker evidence but no clinical phenotype, Dubois gives a diagnosis of “presymptomatic” AD only to those few who carry a pathogenic mutation in APP, presenilin, or have the Down’s syndrome APP triplication. The much larger group of cognitively normal people who have the AD biomarker component but no such gene are told they are “asymptomatic-at-risk” for AD. Dubois acknowledged that data from longitudinal studies is increasingly suggesting that these people have AD and will develop symptoms, but so far he declines to disclose this to individual patients so long as he cannot predict when a given person will progress.
Dubois conceded a point of debate between him and other clinicians. He acknowledged that other episodic memory tests besides the free and cued selective reminding test (FCSRT) also detect the amnestic hippocampal deficit that anchors the clinical part of the AD diagnosis. In addition, the FCSRT may not be sufficiently specific when used by itself. This became clear when Philip Scheltens of Vrije University Medical Center in Amsterdam presented the way screening and enrollment had played out for Roche’s Phase 3 SCarlet RoAD trial of the therapeutic antibody gantenerumab. This trial uses the 2007 Dubois criteria. Among 2,290 potential trial participants who failed screening were 621 who met the FCSRT cutoff but did not have AD biomarkers.
Overall, these criteria drew wide praise for their simplicity. For example, Steve Ferris of New York University said, “I applaud the evolution of these criteria.” “They are very well done,” agreed Charles DeCarli of the University of California, Irvine. Three sets of revised diagnostic criteria have been developed independently in the United States, as well, and are broadly similar.
Even so, the question of how best to apply new diagnostic criteria in clinical trials is far from settled. Different trials are experimenting with slightly different ways of creating groups. Drawing an analogy with infectious disease, Doody suggested that clinical trials might simply divide people into categories of “At Risk,” and define the risk, and “AD,” and define the severity (see image).
At CTAD, Rachelle Doody cited the classification of the Ebola virus to make her point that clinical trials in Alzheimer's could use a simpler approach to creating study populations. [Images courtesy of R. Doody.]
“The most exciting progress we are now implementing in trials is thanks, to a large extent, to new diagnostic criteria,” said Niels Andreasen, who oversees clinical trials at the Karolinska Institute in Stockholm. These research criteria are used primarily in trials and expert settings. For them to become widely used, lumbar punctures or PET scans would need to become more available, and both biomarker measures and memory tests need robust cutoffs. Countries differ in how quickly they embrace the new criteria. In Sweden, many doctors already do, and they offer current AD drugs to barely symptomatic people. “I believe it is unethical to start a patient with biomarkers but no dementia on the placebo arm of a therapeutic trial without offering currently available medicines. If I know from my assessment that this is AD, even if the MMSE [Mini Mental State Examination score] is 28, I want to offer those medications,” said Andreasen.
On the other hand, some clinicians prefer not to subjugate their clinical impression to a biomarker result. For example, the Dubois criteria require either amyloid PET or CSF Aβ/tau positivity for an AD diagnosis, and the regulatory label for amyloid PET tracers states that a negative scan rules out a diagnosis of Alzheimer’s. Even so, in a study at memory clinics in northern Italy, most dementia specialists resisted changing a clinical AD diagnosis they had made in the face of a subsequent negative amyloid PET scan, Giovanni Frisoni of the University of Geneva reported at CTAD. “Dementia experts still rely on the clinical phenotype, even when amyloid PET tells them otherwise,” Frisoni said. Others later quipped that clinicians do not like being shown that they were wrong, but that this will change in the next few years once amyloid and tau PET become more established.
While secondary prevention dominates the discourse these days, there is fresh energy in other areas, too. CTAD showcased a new push to develop hormone-based therapies based on women’s health research (see Part 8 of this series). There was broad excitement about a flurry of trials to try to calm agitation and aggression in Alzheimer’s and even Down’s syndrome (see Part 10 of this series). “Nothing is more important in symptomatic AD than managing the behavioral abnormalities. If this is true, it is big news,” said Aisen.
As new trials get going, studies in symptomatic patients would be well advised to take into account a feature of Alzheimer’s that can stump any longitudinal study, namely that AD progresses faster in some people than others. Both intrinsic and external factors can influence how quickly the disease marches on in a given person. However, new trials could account for different progression rates, said Doody. She suggested collecting the National Adult Reading Test (AMNART) at screening as a surrogate of cognitive reserve and information on background medications, both of which can slow a person’s progression. Parkinsonism and psychosis should be documented, as they can speed up progression. Slow and fast progressors could be balanced across treatment arms, and pre-progression indicators incorporated in the analysis, Doody said.—Gabrielle Strobel
- Rusty Unleashed: Forget Disease Modification, Go for Big Effect
- Large Studies Agree: Brain Amyloid Accelerates Cognitive Decline
- From Shared CAP, Secondary Prevention Trials Are Off and Running
- Immunotherapy I: Baby Steps, but No Breakthroughs
- Immunotherapy II: Active Approaches Down, New Passive Crops Up
- New Target Has Legs: Tau PET, Mice, and Antibodies
- Try This at Home: Cognitive Testing in the Age of Prevention Trials?
- Just for Her? Study of Women’s Biology Offers New Therapeutic Angle
- New Treatments for Alzheimer’s Behavioral Symptoms on Horizon
- Dubois B, Feldman HH, Jacova C, Hampel H, Molinuevo JL, Blennow K, DeKosky ST, Gauthier S, Selkoe D, Bateman R, Cappa S, Crutch S, Engelborghs S, Frisoni GB, Fox NC, Galasko D, Habert MO, Jicha GA, Nordberg A, Pasquier F, Rabinovici G, Robert P, Rowe C, Salloway S, Sarazin M, Epelbaum S, de Souza LC, Vellas B, Visser PJ, Schneider L, Stern Y, Scheltens P, Cummings JL. Advancing research diagnostic criteria for Alzheimer's disease: the IWG-2 criteria. Lancet Neurol. 2014 Jun;13(6):614-29. PubMed.
No Available Further Reading