Neuroscience: Part II

Alzheimer's Disease: The tangled challenge

In Part II of Signals' look at neuroscience, we examine Alzheimer's Disease and the slow and controversial unfolding of science's understanding of this condition's biology. The technical and business challenges of making progress against Alzheimer's are nearly as complex as the neurofibrillary tangles that clog Alzheimer's patients' brains. High-profile researchers still fiercely debate fundamental concepts. And new evidence shows you may add a predisposition to Alzheimer's to the list of grievances you have against your mother. It's more proof that the best-laid plans of mice and venture capitalists can face considerable delay if a concept gets too far out in front of the science. And yet, there is considerable cause for hope. For Neuroscience: Part I, click here.

By Bruce Goldman
Special to Signals

Alzheimer's Disease is a progressive neurodegenerative syndrome that scuttles patients' memories and leaves them demented and often miserable. It represents one of the biggest potential markets in the pharmaceutical industry: Rare before age 50, Alzheimer's afflicts nearly half of all people past the age of 85 -- the most rapidly growing portion of the population. Over 4 million living Americans have been diagnosed with Alzheimer's disease and, barring scientific progress, about 14 million will have it by the middle of the next century. Clearly, any drug that could truly alter the syndrome's course would be an instant multi-billion-dollar hit.

For at least a decade, however, untangling the biology of Alzheimer's Disease has been slippery going despite the concerted efforts of a handful of biotechnology and large pharmaceutical companies. When he joined Athena Neurosciences, a neurobiology company formed in 1986 explicitly to conquer Alzheimer's, recalls Chief Executive John Groom, "The feeling was all the work would be done in about four years." The "work" was the first meaningful therapeutic advance in this disease. But those optimistic early days were "$70 million in research spending ago. Only now are we beginning to see the fruits of that research pay off, and even now the payoff is still several years out," explains Groom.

Indeed, the biological understanding needed to make headway in this condition just wasn't there. The fact that legal and regulatory constraints are especially enervating when it comes to drugs meant for long-term prevention hasn't helped, either. But the science is catching up and the action is picking up. In the 1990s, answers to some basic questions about what causes Alzheimer's have provided biotech companies and drug houses alike with traction. Most scientists are agreed on a hypothesis regarding the cause of the disease, even though it is still the subject of bitter scientific and competitive feuding. New, living lab "instruments" in the form of cell cultures and transgenic mouse models to mimic the disease are speeding drug development. Companies are branching out along increasingly divergent lines of exploration. Several different kinds of Alzheimer's drugs are now on the runway.

A half-true start

In the mid-1970s, Alzheimer's was blamed on a brain deficit of acetylcholine, a neurotransmitter considered to play a primary role in short-term memory. Pharmaceutical companies started beating the bushes for compounds that would enhance acetylcholine stores. But by the mid-1980s it was obvious that plenty of other neurotransmitters were involved and that massive cholinergic nerve-cell loss was just an end-stage event in a decades-long trajectory of deterioration. Any drugs that simply enhanced acetylcholine supplies were doomed to be mere palliatives, setting back atrophy's ticking clock by six months or a year at best.

Still the drug companies pursued drugs that raised acetylcholine levels. It was the only game in town. No deeper understanding of the disease's mechanisms had yet provided a platform for innovative drug development, and demographics were crying out for a drug. The first to come on the market was Parke-Davis's Cognex in 1993. Parke-Davis's parent, Warner Lambert, acknowledges that sales of Cognex, which peaked at $65 million in 1996, are wilting under competition from a similar drug, Pfizer-Eisai's Aricept, launched just a year ago. For one thing, Cognex can prompt serious liver toxicities. For the first 9 months of 1997, worldwide sales of Aricept (mostly in the United States) came in at $128 million. To date, those two drugs are the only approved agents in the Alzheimer's arsenal.

All told, though, close to half a dozen more acetylcholine-related drugs are rumbling down the development pike now, mainly from large pharmaceutical companies. One biotech player, SIBIA Neurosciences Inc. of La Jolla, Calif., will soon begin human trials with a drug dubbed 1553, which in animals has increased acetylcholine levels up to 25-fold in certain parts of the brain, versus two-fold increases for agents such as Aricept. Another, Shire Pharmaceuticals, in a marketing collaboration with Janssen, is into Phase III for its drug.

Everyone realizes these drugs aren't the answer. But some Alzheimer's researchers go a step further and contend these efforts have wasted time and money. Many managed-care organizations don't include even the two FDA-approved brands in their formularies. Dennis Selkoe, a Harvard medical professor and protein chemist, "If I were a pharmaceutical company executive I'd be thinking that for the hundreds of millions of dollars I saw my company waste on cholinergics, there's not much coming from it."

The tragic cascade

Image from University of Oklahoma's web page.

Selkoe, of course, is hardly objective on the matter. In 1986, a group of venture capitalists including Venrock, Domain, and Kleiner Perkins Caufield & Byers, decided to start a neuroscience-oriented biotechnology company. In the interest of jump-starting an Alzheimer's program, they recruited Selkoe as a co-founder of Athena Neurosciences. (Athena is now a wholly owned subsidiary of Elan, based in Ireland). The chief advocate of a hot new "amyloid-cascade hypothesis," Selkoe held that the primary lesion in Alzheimer's was an aggregation, over a period of decades, of dense extracellular gumballs called "senile plaques" that accumulate in Alzheimer's patients' brains and -- along with dementia and the presence of peculiar piles of intracellular debris called "neurofibrillary tangles" -- define the disease.

Image from Eli lilly's web page on Alzheimer's Disease

It was finding these unusual deposits in patients' brains in 1907 that first led German psychiatrist Alois Alzheimer to describe the cognitive and physiological abnormalities that came to be known as Alzheimer's. However, while neurofibrillary tangles track closely with cell damage and death, plaques are more mysterious: They often appear in parts of the brain that seem otherwise unaffected, and, conversely, sometimes are not seen in clearly diseased regions. But because plaques' appearance typically precedes that of tangles by many years, Selkoe and most other researchers believe it's plausible that the build-up of plaque in the brain is responsible for the wholesale nerve-cell destruction that follows.

According to Selkoe's "amyloid-cascade" model, Alzheimer's disease begins when "beta-amyloid," a peptide secreted by brain cells for reasons unknown, begins to aggregate along with certain other proteins into senile plaques. As these aggregates grow denser and more numerous, they eventually enrage nearby glial cells which supply neurons with nutrients and structural support. Inflammatory factors released by angered glial cells damage neurons, whose Lego-like lattices of microtubules are ultimately transformed into a smoldering wreckage of neurofibrillary tangles.

Image from University of Oklahoma's web page.

It's a pretty violent, albeit microscopically violent, picture that Selkoe paints today. But Selkoe's hypothesis lacked empirical ammunition back in 1986 and still awaits definitive proof. Plaque-caused diseases of peripheral organs are well known, says Ivan Lieberburg, Athena's vice president for research. Lieberburg notes that such peripheral diseases can be reversed by reducing the amyloid burden, so it's reasonable to ask whether clearing excess amyloid in the brain might also alleviate Alzheimer's.

Alas, says Lieberburg, the early exuberance of Athena's founding VCs, fueled as it was by the swift ascent of the likes of Genentech and Amgen and their gut feeling that "the time was ripe for a neuroscience-based biotechnology company," was countered by a dearth of hard scientific knowledge. There were no animal models, no decent assays for the disease, no genetic mutations that could point at likely culprits at the molecular level. However, in the ensuing years, the pace of discovery -- both within and without Athena -- has been explosive.

beta-testing, genes yield new clues

It turns out that beta-amyloid is secreted by cells after being chopped from a longer, cell-membrane-bound precursor protein. Since 1991, several mutations in the gene encoding this precursor have been discovered that invariably trigger beta-amyloid overproduction -- along with full-blown, early-onset, familial Alzheimer's disease. In 1995 and 1996, mutations in two other genes, both seemingly unrelated to the gene encoding the precursor protein, were likewise found to cause Alzheimer's, as well as big amyloid build-ups in the brain. The correlation appears to implicate beta-amyloid in the Alzheimer's etiology.

In 1992 Selkoe's lab and others showed that beta-amyloid is secreted by all healthy cells, which ubiquitously harbor the peptide's precursor on their outer membranes. Thus, rather than being toxic per se, amyloid likely is similar to cholesterol: harmful only when the balance between production and clearance is upset. A practical implication: Easily cultured cells can be used to assay drugs for efficacy in curbing beta-amyloid secretion while checking the drugs for cytotoxicity. In 1995, Athena acquired a transgenic mouse containing a mutant human beta-amyloid precursor gene that produces a huge amount of amyloid plaque. In theory, Lieberburg says, it's the best drug screen of all. No cell culture can give you any information on plaque aggregation or behavioral pathology, but a mouse can.

The chicken or the egg or the Apo E?

A few prominent researchers think the emphasis on amyloid is misplaced, however. Selkoe and Duke University geneticist Allen Roses are embroiled in a running debate about that and lots of money is riding on the outcome. The Selkoe-Roses debate hinges on causality: Does beta-amyloid actually cause the disease? Or are plaques mere byproducts of Alzheimer's neurodegeneration? "Competition between theories is great for academia," says Roses, "but it can be expensive for companies."

In 1993 Roses and colleagues at Duke knocked the Alzheimer's research community for a loop by implicating an unlikely suspect -- apolipoprotein E (nicknamed "apo E"), a lipid-shuttling substance that until then had been widely studied mainly for its involvement in atherosclerosis. For his labors, Roses attracted considerable criticism as well as delays in publication in scientific journals and niggardly funding by granting agencies, he contends. But then Zaven Khachaturian, who for many years as director the National Institute of Aging's Office of Alzheimer's Disease Research oversaw allocation of the overwhelming majority of the federal government's Alzheimer's research funds, started taking Roses seriously and channeled funds his way. "Apo E comes in three flavors," explains Khachaturian. "Everybody gets two scoops." Numerous large population studies have now firmly established (although nobody's sure why) that the three different human variations, or alleles, of the gene coding for Apo E confer different likelihoods of developing Alzheimer's, as Duke's Roses first asserted. Of equal significance for those who will eventually succumb to the disease, inheriting different pairs of Apo E-coding alleles can mean a spread of as much as two decades in the average age at which symptoms begin to manifest.

This summer, Roses became vice president and worldwide director for genetics at Glaxo-Wellcome, where he says he now gets more research funding in one year than in his entire history as an academic. He plans to use that money to bring a whole new category of Alzheimer's drugs to market. Roses contends that Alzheimer's disease is simply a natural outcome of the aging process within the living brain, a process whose rate is somehow influenced by Apo E. "We would all get Alzheimer's if we all lived to 140," he maintains. While declining to discuss specific therapeutic-intervention targets, Roses does suggest that a small molecule that caused the high-risk Apo E variety to behave like the low-risk one would in effect postpone the age of onset by a decade or two, to well beyond our average life expectancy. That would in turn markedly lower the prevalence of Alzheimer's among older people.

This medical miracle, should it occur, would owe nothing to the now prevailing amyloid-cascade hypothesis. "Most companies that claim to be doing Alzheimer's research have actually been doing amyloid research," says Roses, who has repeatedly professed skepticism about the peptide's pivotal etiological role.

The amyloid buffs rebuff Roses' agnosticism. Contends William Comer, CEO of SIBIA, "Ninety-eight percent of the gurus believe it is beta-amyloid deposits that cause Alzheimer's." A recent crossing of the Athena-Lilly amyloid-hypersecreting mouse with another animal incapable of producing apo E yielded a hybrid that, while still producing excess amyloid, had greatly diminished plaque formation -- what Lieberburg calls incontrovertible confirmation of an apo E-amyloid link, suggesting that Apo E's connection to Alzheimer's may be as an escort service for beta-amyloid. Indeed, Lieberburg, Selkoe, Khachaturian, and others all note that what originally drew Roses' interest to Apo E as having a possible connection to Alzheimer's was the discovery, in the lab of his Duke colleagues, that Apo E has a strong binding affinity for none other than beta-amyloid.

"It walks like a duck, it quacks like a duck," shrugs Lieberburg. "Isn't it reasonable to suggest that it may, in fact, be the duck?"

Closing in on the answers

Is amyloid the culprit or not? In a few years we may know.

In addition to its cholinergic work, SIBIA is working with Bristol-Myers Squibb on a compound that eliminates the formation of beta-amyloid in 80-90 percent of mice tested. They hope to begin human trials soon. Athena and its partner Eli Lilly have run hundreds of thousands of compounds through Athena's plaque-ridden transgenic mouse. Some of these compounds lower beta-amyloid production and the collaborators hope to move into clinicals before too long, as well. Assuming such compounds do lower human subjects' beta-amyloid levels, it could take 18 to 24 months to see if the rate of the subjects' functional deterioration slows, too. "That's the proof of the pudding," acknowledges Lieberburg. "The mouse is not going to tell us whether we have a cure for Alzheimer's. The true test is going to be, What does it do in people? Is treatment with the drug going to slow the progression of dementia?" Roses waits with gently smiling jaws. "If a drug that lowers amyloid doesn't decrease dementia," he says, "then we won't be talking Athena's current place in research -- we'll be talking about it in the past tense." And if it does decrease dementia? "If we really thought they were right, we probably would have bought them."

While some believe the above studies will definitively solve this puzzle, others believe the answers are unlikely to be clear-cut, and that conquering Alzheimer's ultimately will require different, layered strategies. "I don't think the ultimate market is going to be made or broken by one drug," says Khachaturian, now an independent consultant who is starting his own company. "The nervous system is very complex. We're going to see different compounds being used at different stages of the disease. And there will still be room for old-fashioned replacement therapy; some newer compounds may extend the useful life of the older drugs."

While neither amyloid nor Apo E is likely to go away, some companies are hedging their bets by adopting a multifaceted approach. Athena, for example, has an active anti-inflammatory program. Nobody claims it's the primary event, but everyone agrees that inflammatory mechanisms play a key role. Evidence of inflammatory involvement has been mounting since a study of rheumatoid arthritis patients on anti-inflammatories a few years back showed that they had a lower-than-chance incidence of Alzheimer's. Several big National Institutes of Health (NIH)-funded trials of prednisone, a synthetic form of the glucocorticoid cortisol, are underway now, and every big drug company with an arthritis program is watching.

Another area of great interest is in neuroprotective compounds. Even in the face of major damage, neurons have a surprising capacity to regenerate if they are supplied with certain growth factors. But these are typically large proteins that don't cross the blood-brain barrier. Plus, Alzheimer's involves perhaps a dozen different classes of dying neurons, each possibly requiring its own special trophic factor, and cell death is distributed throughout the cortex and limbic system.

Athena's Lieberburg derides the exotic delivery systems some companies envision to deliver these factors, such as implanted pumps or CytoTherapeutics' technique of embedding transfected cells in a gel mesh so they can squirt out factors safe from immune cells' reach. "What are you gonna do?" Lieberburg asks. "Insert these things in an array all over your head? That's major surgery. I don't think managed-care organizations would look favorably on that. And I doubt very much that it would work. Wouldn't it be better to just make a pill?"

Take a pill?

Cephalon is doing just that, says its vice president for research, Jeff Vaught. One of two promising pills, both of which the company expects to see in clinical trials by sometime next year, is a vitamin D analog that turns on nerve cells' neurotrophin genes. The second compound, a therapeutic mirror image of the first, acts by shutting down the cell-death pathway. Several companies, including NeoTherapeutics, American Biogenetic Sciences and Neurocrine, have announced progress with small molecules that appear to cross the blood-brain barrier, but behave like neurotrophic factors once inside.

Acting on the notion that inflammation triggers oxidative damage to nerve cells, Centaur Pharmaceuticals is collaborating with Astra AB on an anti-oxidant compound. Cortex Pharmaceuticals has had promising Phase I results with a compound that, by enhancing the action of glutamate, the brain's chief excitatory neurotransmitter, appears to improve patients' short-term recall; a Cortex official speculates that because nerve cells live longer when they remain active, the compound may even prove to slow nerve-cell degeneration. Unimed is also in Phase I, but with a marijuana derivative that, while in no way altering the course of Alzheimer's neuropathology, may alter patients' behaviors in ways that will allow caregivers to keep them at home longer.

Last year, more basic revelations of the biology of Alzheimer's opened up new research directions, as well as helped explain the results of another study showing that vitamin E, an anti-oxidant, can slow progression of Alzheimer's. Scientists at San Diego's Mitokor showed that mutations in mitochondrial DNA, which are maternally inherited, may contribute to a defect in brain cells' energy metabolism. The defects appear to occur in genes that convert glucose and oxygen into energy -- instead producing so-called free radical molecules that are destructive to cells. The loss of energy and the damage from free radicals may allow beta-amyloid to build up, researchers theorize. And since anti-oxidants scavenge free-radicals, that would explain the Vitamin E findings. Mitokor is currently making tests for the mtDNA mutations available to companies doing clinical trials in Alzheimer's.

The challenge ahead

In an ironic twist, Athena -- the company that Selkoe founded --holds the license to a clinical apo E-allele assay Roses designed at Duke. Roses says the test is 95 percent accurate in the presence of dementia. However, the ultimate goal, of course, is not to identify Alzheimer's after the fact, but to start treatment early so that symptoms come at age 90 or 100 instead of age 60 or 70. Progress on that front may benefit from a few regulatory paradigm shifts. When Merck tested Mevacor, the FDA approved it on the basis of its lowering cholesterol, before any reduction in heart attacks or total deaths could be shown. If beta-amyloid, for example, proves to be to Alzheimer's what cholesterol is to cardiovascular disease, rendering anti-amyloid drugs analogous to statin compounds, FDA acceptance of surrogate markers for Alzheimer's deterioration, such as lowered levels of circulating beta-amyloid, would speed trials.

Patent protection is another thorny issue. "It can take 15 years to show a lowered Alzheimer's incidence in a prevention trial. By the time the product is brought to market, there may be only one or two years left on the patent." says Khachaturian. However, Comer of SIBIA says surrogate markers could help solve that problem: "I believe there will be drugs approved for lowering beta-amyloid for several years before we have a label that says we're curing Alzheimer's."

Above all, early treatment will require early diagnosis. Accurate early assays are undoubtedly going to become a big profit center. Managed care organizations will want to see patients properly diagnosed before okaying costly long-term preventive treatment. If early diagnosis is important to drug providers, it is even more critical to patients. "There's no point in doing anything after a late diagnosis," says Khachaturian. "All you're doing is expanding the duration of the disability -- unless somebody comes up with a molecule that actually reverses the course of the disease, and that's way in the future. Even then, it won't restore memories. What's lost is lost."

Let's hope this is the last generation of patients who will feel that loss so early . . . and so harshly.