Alzheimer's Immunotherapies: Reasons for Hope


by Jim Schnabel

June 20, 2011

The idea of immunotherapy against Alzheimer’s began promisingly in the late 1990s, with the demonstration that an active vaccine targeting the protein amyloid-beta (A-beta) could clear or prevent A-beta brain plaques in a mouse model of Alzheimer’s. But disappointments followed. The first effectiveness study of this vaccine in humans had to be halted prematurely in early 2002 after some subjects began to suffer life-threatening brain inflammation. There was further disappointment in 2008 when another clinical trial, of a safer, passive immunotherapy that simply infuses anti-A-beta antibodies into patients, showed little sign of effectiveness.

The Alzheimer’s immunotherapy story isn’t over, however. Large-scale trials of several active and passive immunotherapies are nearing completion, and there are reasons to hope that one or more of these will show at least a modest benefit for people with established dementia. If that happens, immunotherapies will probably end up being targeted more precisely against disease-associated forms of A-beta, and would also be used earlier and earlier in the disease process, “with better therapeutic outcomes,” says Thomas Wisniewski, a neurologist and immunotherapy researcher at New York University’s Langone Medical Center.

Active vs. passive

In principle, an active vaccine needs to be injected only once or a few times to stimulate the immune system to ward off disease indefinitely. This could be the ideal way to prevent Alzheimer’s and other diseases that are driven by amyloid-forming proteins. Because only a few doses would be needed, such a vaccine could be relatively inexpensive, and eventually could be be given to people when they are still young and healthy.

However, testing an active vaccine in a true disease-prevention role would take decades, and could only be done with a vaccine that is clearly safe enough to use in healthy people. Testing an active vaccine as a therapy for existing dementia takes less time and is ethically more justifiable. In fact, this testing is happening now, with safer successors to the original encephalitis-causing “AN-1792” vaccine. But here again there is a problem:  Elderly people with their weaker immune systems respond less well to active vaccines. In the ill-fated AN-1792 trial, for example, only 19 percent of treated patients made what researchers considered adequate levels of antibodies against A-beta—and a large proportion of those also developed autoimmune encephalitis.

Passive immunotherapies are meant to get around this problem by delivering anti-A-beta antibodies in an intravenous infusion every few weeks or months. Many researchers see these antibody infusions as being the best hope for elderly people with weakened immunity, whereas the active vaccines—if they could show sufficient safety and efficacy in a trial—would be the best solution for younger people with healthier immune systems. “That’s pretty much the scenario I’m envisaging if all this pans out,” says Wisniewski.

Unfortunately, the first passive immunotherapy to be developed, a solution of monoclonal (single-type) anti-A-beta antibodies known as bapineuzumab, showed a hint of effectiveness in only a subset of Alzheimer’s patients in its initial large-scale trial—despite brain-imaging evidence that the antibodies markedly reduced patients’ A-beta plaques. Along with several competing products, it is now in larger-scale trials due to finish next year. But even if it works in some patients, it—and probably its competitors, too—has safety issues. The antibodies in passive immunotherapies bind very strongly to A-beta, and are infused over an hour or two, so that they appear in the blood much more quickly than they would after active vaccination. Results from bapineuzumab’s trials show that brain swelling and associated microbleeds from cerebral blood vessels can occur, especially in patients with heavy A-beta deposits in those vessels—a condition that isn’t easy to detect with current diagnostic methods. “My guess is that it has something to do with the removal of A-beta from the blood vessel walls by the bolus of antibodies,” says Cynthia Lemere, a Harvard Medical School researcher who has been working on A-beta vaccines since the late 1990s.

New blood?

The Alzheimer’s immunotherapy with the greatest near-term promise is neither an active vaccine nor a solution of A-beta-targeting monoclonal antibodies. It is an intravenous solution containing a broad mix of antibodies sifted from the blood of healthy human volunteers. Known as intravenous immune globulin or “IVIg,” it is made by several drug companies around the world, isn’t patentable, and has been used for years to treat other immune-related conditions.

Preliminary trials of IVIg, overseen by neurologists at Cornell University’s Weill Medical College, have gone surprisingly well in Alzheimer’s patients so far. Treated patients’ cognitive test scores have usually improved moderately and then stabilized, while the placebo group’s scores have worsened as expected. The most recent trial included only 24 patients, and such small studies often show promising results that fail to translate to larger trials. However, in this case, the researchers saw evidence of benefits for IVIg-treated patients by several other “biomarker” measures including A-beta levels in cerebrospinal fluid, positron-emission-tomography imaging of brain metabolism, and magnetic resonance imaging of brain volumes.

“I would say that IVIg is the leading anti-amyloid strategy right now,” says Sam Gandy, an Alzheimer’s researcher at Mount Sinai School of Medicine.

The National Institute on Aging (NIA) and Baxter Healthcare, which makes a commonly used IVIg product, are now supporting a much larger multi-center trial which will enroll about 400 people and run for 18 months of treatment. It is currently due to finish late in 2012 or early in 2013. If successful, it would almost certainly lead to FDA approval of IVIg’s use in Alzheimer’s, and if its benefits last the full 18 months, it also would be acknowledged as a “disease modifying” therapy. No such therapy exists at present. However, since IVIg is on the market now for immunological conditions, some neurologists already prescribe it “off label” for their Alzheimer’s patients.

Looking ahead

Passive immunotherapies such as IVIg and bapineuzumab may end up helping many people, but even if they do, Alzheimer’s patients and their families—and their insurers—would soon want better strategies. Patented monoclonal products would be very expensive and would have to be given every few weeks or months, perhaps indefinitely. Even IVIg is expensive; each dose used in its current trial would cost a least several thousand dollars. Worse, because it can’t currently be produced synthetically, but must be sifted from human donor blood, it may never be produced in quantities that could satisfy the demand from Alzheimer’s patients and their families. Thus its very limited supply would be another factor restricting its use.

On the other hand, even a modest success for one or more immunotherapies in ongoing trials should provide clues to the making of better therapies and preventives. Researchers now believe that small clusters of A-beta, known as oligomers, are much more toxic to the brain than are the larger fibrils that make up plaques. A-beta oligomers also appear to promote the appearance of deformed tau proteins, whose clustering and spread late in the disease corresponds closely to the worsening of dementia.

“The ‘secret ingredient’ of IVIg is believed to be [natural] anti-A-beta oligomer antibodies,” says Gandy, who recently led the development of an A-beta-oligomer mouse model of Alzheimer’s. “Any effects on tau might be downstream of the effects on oligomers, although I don’t know whether anyone has actually looked in IVIg for anti-tau antibodies.”

If multiple protein oligomers drive Alzheimer’s dementia, then the ultimate strategy could be a non-protein-specific anti-oligomer vaccine, which attacks the special toxic shape that disease-causing oligomers such as A-beta and tau appear to have in common. Wisniewski is one of several investigators now developing these. In a study published last October, his lab’s active anti-oligomer vaccine elicited antibodies that recognized both A-beta and tau aggregates, and restored the behavior of Alzheimer’s mice to near normal. Yet these antibodies apparently didn’t attack A-beta fibrils, such as those found in vascular deposits, as fiercely as bapineuzumab and other monoclonal antibodies would. “That suggests that [the vaccine] is less likely to be associated with hemorrhages and encephalitis,” he says.