UD professor seeks Alzheimer’s tool and treatment
Kelvin H. Lee, Gore Professor of Chemical Engineering and Delaware Biotechnology Institute Faculty Fellow, is hoping that the work he and his research group are doing will contribute to the development of both an accurate diagnostic tool for Alzheimer’s and a strategy that will protect against the ravages of the disease.
1:34 p.m., April 29, 2008--Alzheimer's disease (AD), the most prevalent form of dementia, affects some 4 million people in the United States. It is the third most costly disease, after cancer and cardiovascular disease. And not only is there no cure for Alzheimer's; there is no accurate pre-mortem diagnostic tool. Only an autopsy can indicate for sure that a patient had the disease.

Kelvin H. Lee, Gore Professor of Chemical Engineering and Delaware Biotechnology Institute (DBI) Faculty Fellow, is hoping that the work he and his research group are doing will contribute to the development of both an accurate diagnostic tool for Alzheimer's and a strategy that will protect against the ravages of the disease.

Lee came to the University of Delaware in September, after spending 10 years on the faculty at Cornell University. He holds a doctorate in chemical engineering, with a minor in biology, from the California Institute of Technology. The minor in biology reflects Lee's interest in the life sciences and his focus on medical applications of chemical engineering research.

His vision is grand--he would like to contribute to the development of a tool like the medical “tricorder” seen in Star Trek. The handheld device could scan the body, interpret and display data from the scans, and record information, helping doctors to diagnose disease. Given the rapid development of proteomics, the large-scale study of the structure and functions of proteins, Lee's vision may not be as far-fetched as it seems.

“We're looking for changes in protein expression in cerebrospinal fluid,” he explains, “and trying to come up with a 'bar code' that can distinguish between patients who have AD and those who don't, as well as between people with AD and those with other, similar diseases that can cause dementia.”

Doctors can currently diagnose living patients only as “probable AD” because it takes a post-mortem examination of brain tissue to provide definitive evidence of the amyloid plaques that characterize Alzheimer's. Lee points out that an estimated 10-20 percent of people with this diagnosis are found to have other conditions that manifest similar symptoms.

“We need a tool that's specific enough to distinguish among neurodegenerative diseases so that the proper treatment can be administered,” he says.

As a postdoc at Caltech, Lee had worked with a research group that was studying Creutzfeldt-Jakob disease (CJD), a rare and incurable brain disorder that, like Alzheimer's, can be diagnosed only post-mortem through an examination of brain tissue. CJD is classified as a transmissible spongiform encephalopathy, a term used to designate a group of disorders that also includes mad cow disease. Using a special technology to measure “protein fingerprints” in cerebrospinal fluid, the researchers found that there was a change in one particular protein in people who had CJD.

The project provided a proof of principle that protein fingerprints could be used as a tool to diagnose neurodegenerative disease, but when Lee accepted a faculty position at Cornell in 1997, he wanted to shift his focus. “Mad cow was biologically interesting and was making headlines at the time because of an outbreak in England,” he says, “but it's not a disease that affects a lot of people.”

So Lee decided to focus on Alzheimer's. Working with Dr. Norman Relkin, a physician at the Cornell Medical Center in New York City who was doing both clinical and laboratory work on the disease, Lee received funding from the National Institutes of Health for a study aimed at finding biomarkers for the disease--in effect, a protein barcode unique to those with Alzheimer's.

The results of the study yielded a set of validated biomarkers for Alzheimer's, which was published in the Annals of Neurology in late 2006. In the meantime, the research group had turned its attention from diagnosis to treatment.

“There has never been a good treatment for Alzheimer's,” Lee says. “Most therapies treat the symptoms to improve quality of life for six months or so. At the end of that period, many patients are not any better off than untreated patients.

“My collaborators began to wonder whether people could be immunized against AD,” Lee continues. “The disease is characterized by the formation of amyloid plaques in the brain. Would it be possible to get the body to form antibodies to clear the damaging plaques when they're formed?”

Relkin and colleagues from Cornell conducted a Phase I study in which they administered intravenous immunoglobulin (IVIg) to Alzheimer's patients. “The drug was already FDA approved for treating certain people with compromised immune systems,” Lee says.

He explains that IVIg is basically a cocktail of antibodies taken from a pool of healthy donors. “It turns out that we normally produce antibodies to the protein that is a hallmark of Alzheimer's,” Lee says.

The Phase I study design involved six months of treatment with the drug, followed by three months of no treatment, referred to as drug washout.

“Cognitive ability improved during the treatment period and then reverted during drug washout,” Lee says, “suggesting that cognitive losses were actually reversed by the therapy. This was a very encouraging outcome despite the fact that it was preliminary and involved a very small number of subjects. As a result, all of the subjects were put back on the drug, which resulted in positive effects in patients even after 18 months.”

Lee points out that the Alzheimer's research conducted by his team began in one direction and has changed over time. The initial effort, which has basically been accomplished, was to identify diagnostic markers for Alzheimer's and to determine what happens to the markers as patients are treated.

“We found that there are markers and that they mirror the clinical results,” Lee says. “In other words, improvement in patients' cognitive functioning seemed to correlate with a reduction of the markers in their cerebrospinal fluid.”

The next steps for the team include conducting additional clinical trials and moving towards obtaining FDA approval to treat Alzheimer's patients with IVIg.

“To facilitate this, we would like evidence that the clinical outcomes observed are derived from measurable molecular changes,” Lee explains. “We want to demonstrate, at the molecular level, the disease-modifying effects of the drug.”

Lee is still collaborating with his Cornell colleagues to collect data that will validate the link between the disease-modifying effects of the drug and the observed clinical outcomes. He said he hopes that their work will lead not only to a noninvasive diagnostic tool and an effective treatment but also an understanding of the mechanisms underlying the development of the disease.

“It would be ideal to have that medical tricorder,” he says. “But we're not there yet.”

Article by Diane S. Kukich
Photo by Kathy Atkinson