One recent afternoon, in a newly renovated 12th-floor research complex in Columbia University’s College of Physicians and Surgeons, Alzheimer’s researcher Karen Duff inspected several elderly mice sitting in cages on a shelf in her lab. These were, in lab parlance, tau mice. They had been genetically engineered to produce abnormal human tau protein in a very specific part of their brains, the same small place where autopsies have shown that it first appears in human brains. One mouse in particular stood out because of its ragged, ruffled brown fur.
“This one may be a bit demented,” Duff said matter-of-factly. “It’s a little less well-groomed, and one of the first signs [of dementia] is a rougher fur.” If these mice mimicked the pattern of pathology seen in humans with dementia, Duff added, the misshapen protein “would have spread to areas of the brain affected by Alzheimer’s disease.” The confirmation came a few days later, when technicians sacrificed the animals and mapped tangled bits of malformed tau that had spread throughout their brains. It is these tangles, according to Duff, that eventually kill the cells that confer memory, perception, cognition.
The mice on Duff’s lab shelf and their experimental brethren have introduced a surprising new wrinkle to the pathology of Alzheimer’s. Duff and her colleagues have conducted experiments showing that misshapen tau proteins initially sequestered in the part of the brain where Alzheimer’s typically first appears (the entorhinal cortex) somehow were able to spread along nerve circuits and hop across synapses to other parts of the brain long known to be involved in dementia, including the hippocampus. As these abnormal tau proteins spread through the brain, they “usurped” and corrupted the normal tau proteins in affected cells, inducing lethal tangles and killing neurons.
The good news is that this mechanism offers novel opportunities for treatment: attacking abnormal tau as it hops between cells. The Columbia group is already conducting animal tests of a monoclonal antibody designed to intercept tau at precisely this vulnerable point of passage, and Duff says pharmaceutical companies have shown considerable interest in the model.
But the new findings are also a stark reminder of how much researchers still need to learn about Alzheimer’s in particular and dementia in general. The scientific literature now describes amyloid as necessary but not sufficient to explain Alzheimer’s symptoms, yet despite intense investigation, there is no general agreement on the mechanism linking the two signal features of a brain in the throes of the illness. Scientists still don’t know why the amyloid plaques precede the tau tangles by anywhere from 10 to 20 years, and they don’t know how the two pathologies are connected. “We know amyloid protein builds up, but there’s a lot of debate if it’s the chicken or the egg, if that’s the trigger or the result of the disease,” says Brangman.
Even after decades of discussion about the role of amyloid in Alzheimer’s disease, researchers concede, the hypothesis that these plaques are key to the illness has not been properly tested. “We haven’t tested the right patients at the right time with the right agents,” says Greenberg. “The reality is that we haven’t done that yet. But the field knows what to do and is doing it now.”
Indeed, several ambitious clinical trials—Greenberg considers them the most important trials in the history of Alzheimer’s drug discovery—are poised to launch in the next few months, and the results will shape dementia research for years to come. If these so-called prevention trials succeed, they will hold out hope that the usually inevitable course of dementia can be altered. If they fail to modify the course of the disease, however, the implications will be what researchers like Greenberg and Duff call “devastating” and “horrendous.”
Given the magnitude and urgency of the problem, it’s no wonder that when Kathleen Sebelius, the U.S. health and human services secretary, announced new NIH funding last February, she told reporters, “We can’t wait to act.” And yet it’s clear to many experts that we probably will wait for an effective Alzheimer’s drug—perhaps as long as 10 or 15 years.
The challenge of finding a treatment that will alter the course of dementia is daunting precisely because the process of neural degradation proceeds invisibly for so many years and starts so early. How early? Last July, the Dominantly Inherited Alzheimer’s Network, a network of leading academic centers based at Washington University in St. Louis, published surprising findings that detectable changes in amyloid chemistry in patients with a genetic form of Alzheimer’s may appear in a person’s cerebral spinal fluid up to 25 years before the onset of Alzheimer’s symptoms. By the time Alzheimer’s patients show up in the neurologist’s office with signs of mild or moderate dementia, it is too late.
If the amyloid hypothesis for Alzheimer’s is correct, therefore, researchers need to find and treat patients a decade or more before the first signs of cognitive impairment appear. They need a drug that crosses the blood-brain barrier to disrupt the buildup of amyloid. And they need diagnostic tools—the cognitive and neural equivalent of a glucose test for diabetics—to measure changes in amyloid and other biomarkers to determine if the therapies are working. (These same diagnostic markers might also be used to identify patients at risk for Alzheimer’s who would benefit from preventive treatment.) Although progress has been made in finding these markers, their reliability is still uncertain. The Food and Drug Administration could speed up drug approvals on the basis of improvements in them, says Sam Gandy, director of the Mount Sinai Center for Cognitive Health in New York. But everyone will still be “holding their breath” until patients are “well beyond the age at which they would be expected to be at risk of becoming demented.”
The group of researchers based at Washington University has assembled a promising tool kit to help them detect the progress of the disease: brain imaging of amyloid deposits, analysis of cerebral spinal fluid, and cognitive tests. But who should the test subjects be? As it turns out, there are several rare genetic forms of Alzheimer’s, and these have been the network’s longtime research focus. People who inherit very specific dominant mutations are fated to develop Alzheimer’s at a relatively early age, and researchers can calculate when the first symptoms of the disease are likely to appear. The network is now in the final stages of selecting three distinct therapeutic agents that target amyloid, with plans to test them in patients with genetic forms of Alzheimer’s.
Randall Bateman, a Washington University physician and researcher, says the aim of the study is to find a drug that will curtail the buildup of amyloid in the brain, much as doctors use statins to reduce the risk of stroke and heart attack by lowering cholesterol levels. Bateman says his research group hopes to launch human tests using the biomarkers by early 2013; he and his colleagues hope to see evidence of effects on these markers after two or three years of treatment rather than waiting 10 or 15 years, when symptoms of dementia would be expected to appear.
The other closely watched trial will be launched—with the NIH’s blessing and funding—by the Banner Alzheimer’s Institute in Phoenix and Genentech. Most of the patients in this trial also have a genetic form of the disease. Members of an extended family of some 5,000 people living in the Antioquia region of Colombia are at risk for a very rare mutation; those who are carriers invariably develop an early-onset version of Alzheimer’s. The idea is to treat about 300 members of this group with an experimental drug to attack amyloid plaques an estimated 15 years before symptoms would be anticipated.
The drug, licensed by Genentech, is an amyloid-attacking monoclonal antibody called crenezumab. Doctors believe it can safely be injected at a higher dose than other monoclonal drugs. “We believe the higher dose will translate into higher efficacy,” says Carole Ho, group medical director for early clinical development at Genentech.
In administering these drugs earlier, to a population genetically susceptible to the disease, Alzheimer’s researchers believe they are finally giving the right kind of therapy to the right patients at the right time. And given the stakes, the two prevention trials have sparked high anticipation. “This will be the first true test of the amyloid hypothesis,” says Barry Greenberg. “The strategy is sound. So let the data happen.”
If the prevention trials succeed, however, there’s no guarantee that this version of early intervention will help in most cases of dementia. Clinicians warn that these rare, early-onset, mutation-based forms of the disease account for at most 10 percent of all Alzheimer’s cases. As Evelyn Granieri puts it, “This may not even be the Alzheimer’s disease that the majority of people get.” The genetic forms of the disease are similar in pathology to the forms most people do get, Ho says. Still, even positive results gleaned from early interim analyses of these trials would come too late for the millions of people who have already begun the slow descent into cognitive decline. “The reality,” says Granieri, “is that most people who are around and sentient now are not going to be around for the cure.” All the more reason, according to Greenberg, to adopt “fundamentally different thinking” in dementia research. “The medical-care system is going to be bankrupt by 2050 if we don’t figure out a way to delay or treat Alzheimer’s disease,” he says, and he believes that won’t happen without a major public-private international initiative. “The competitive marketplace,” he says, “was not conceived to overcome problems of this magnitude.”
Stephen S. Hall’s latest book is Wisdom: From Philosophy to Neuroscience (Vintage). His last story for Technology Review was “The Genome’s Dark Matter.”