Researchers are using lumbar punctures, or spinal taps, to collect cerebrospinal fluid, or CSF, the liquid that bathes the brain and spinal cord. They think chemicals in the fluid may be key to predicting and understanding neurological disease. Some day these chemicals, called biomarkers by scientists, may help predict brain problems, just as high cholesterol can herald trouble in the heart.
Doctors already routinely use CSF to diagnose infections and multiple sclerosis. But in the past, they found few clues there relevant to the memory loss and dementia of Alzheimer's or the tremors of Parkinson's. Now, with newer technologies to identify CSF chemicals and promising results on the power of CSF from a large study of Alzheimer's disease, scientists have renewed interest in spinal taps.
"It allows you to have a window into the biochemistry of the brain," says Dr. Joy Snider, a neurologist at the Washington University School of Medicine in St. Louis.
Scientists collaborating in the Alzheimer's Disease Neuroimaging Initiative, or ADNI, have found that certain CSF proteins go up and others down in people with the disease — and also in people who are likely to get it in the future. Now scientists studying Parkinson's are looking to find similar markers in a project launched this year, funded by the Michael J. Fox Foundation for Parkinson's Research. And in the case of Lou Gehrig's disease, also called amyotrophic lateral sclerosis, scientists have found several CSF changes that might someday help them predict the severity of disease in a patient.
This is a spinal tap: Patients curl forward to stretch apart the bones in the back, and a doctor injects anesthetic to numb the lower back. The doctor then slides a needle into the lower spinal canal, just a bit deeper than anesthesiologists do for an epidural. It takes a few minutes for a bit of the clear, watery fluid to drain into a tube. Afterward, a person usually lies down for half an hour or so.
Although nobody likes the idea of a needle in the spine, a lumbar puncture is fairly safe, Schwarzschild says. "For most people, it's not much different from giving blood."
For diseases like Alzheimer's, by the time a person has trouble with memory or dementia, many brain cells are already dead or dying. By late stages, "it's hard to do much to arrest the disease or rescue the brain from the horrors of this awful disease," says Leslie Shaw, a biomarkers researcher at the University of Pennsylvania Medical Center in Philadelphia.
There is an ongoing "quest," Shaw says, to find ways to catch Alzheimer's early. In 2004, several researchers banded together to start the $70-million ADNI project, hunting for predictive biomarkers in spinal fluid as well as telltale signs of early Alzheimer's in brain scans.
In several studies, ADNI researchers and other scientists have found clear biomarkers in the CSF of people who have Alzheimer's disease. The amount of a protein called amyloid-beta — the ingredient of the plaques that form in the brains of people with Alzheimer's — drops as the disease worsens. Conversely, levels of another protein, called tau, go up. Tau forms abnormal, tangled structures that appear, along with plaques, in the Alzheimer's brain.
The scientists found that this low-amyloid, high-tau signature appears not only in people with advanced Alzheimer's disease but also in people who may be headed that way. In a 2009 paper in the Archives of Neurology, Snider and colleagues studied 49 people with mild thinking problems, which often lead to Alzheimer's. They found that those in the group who advanced most quickly to Alzheimer's were those who had low amyloid and the high tau in their CSF.
And in this month's Archives of Neurology, the ADNI group reported that a person can have this CSF signature of Alzheimer's without having any memory problems at all. They found, in fact, that one-third of healthy elderly people in their sample of 114 had low amyloid and high tau. It could be those people are in the early stages of Alzheimer's, but the scientists are cautious about that interpretation.
"That doesn't mean they're going to get Alzheimer's tomorrow or anything like that," Shaw says. Just as high cholesterol isn't a guarantee of a heart attack, some people with the CSF signature could be headed toward Alzheimer's and some might not. Researchers do not yet have enough information, Shaw says, to use these biomarkers to predict if and when Alzheimer's will arise in healthy people. ADNI will be following those people as long as possible, he adds, to see if they develop Alzheimer's.
The success of ADNI inspired the Michael J. Fox Foundation to launch a similar study, the $40-million Parkinson's Progression Markers Initiative, this year. Over its five-year course, doctors in the U.S. and Europe intend to examine CSF chemicals, among other samples, from 400 people with Parkinson's disease.
Schwarzschild and his colleagues have already discovered one interesting chemical in the CSF of people with Parkinson's: urate, an antioxidant that may protect brain cells. In 2009, the scientists reported that, on average, people with plenty of urate in the CSF and blood tend to have a slower progression of Parkinson's than people with low urate levels.
That does not necessarily mean that urate levels are a direct cause or consequence of Parkinson's, Schwarzschild notes, and it is too soon to use urate alone to predict the course of disease. But he is now treating people with urate supplements in a small clinical trial to see if the treatment is safe and if it boosts urate levels in the blood and CSF of people with Parkinson's. If so, researchers will probably study whether urate supplements slow Parkinson's.
In many cases, biomarkers may be most helpful in predicting severity, not diagnosing disease, says neuroscientist James Connor of the Penn State Hershey Medical Center. For example, neurologists don't need much help diagnosing Lou Gehrig's, he says. But telling those patients what to expect is harder. Many patients die within a few years, but others may survive for a decade or more. "It matters to the patient: Are you on the three-year plan or the 10-year plan?" he says.
Connor is working to find biomarkers that may help doctors make those kinds of predictions. In the journal Neurology, in 2009, he and colleagues reported that 11 proteins related to inflammation go up in the CSF of people with Lou Gehrig's.
Researchers trying to understand disease and develop treatments also find clues in the CSF. For example, knowing that people with Lou Gehrig's disease have inflammation in the spinal cord suggests scientists should consider inflammation as a process to target with drugs. And by tracking changes in the pattern of amyloid and tau in the CSF of people with Alzheimer's, doctors might get a sense of whether an experimental drug treatment is working.
In the future, doctors might even screen elderly people for biomarkers for Alzheimer's or Parkinson's, and treat the disease early just as they do for people with high cholesterol — perhaps staving off or slowing diseases that, for now, are impossible to prevent.
"If we have a treatment that works even a little bit in people with the disease, it may work a lot better in people that do not yet exhibit signs of the disease," Snider says.