A Direct Hit to Parkinson's?

Infusing GDNF shows promise for brain-cell regeneration

For the victims of brain diseases, the golden hope is that scientists will figure out how to make damaged cells deep inside the brain regenerate. If this magic capability could be extended to the frayed neurons of patients suffering from Alzheimer's, stroke, Lou Gehrig's disease, or Parkinson's, their brains might restore themselves, much like skin that heals after a cut.

For Parkinson's sufferers, this dream may be moving toward reality thanks to the revival of a drug written off as a failure just three years ago. A clinical trial at the University of Kentucky is currently treating 10 Parkinson's patients with a bioengineered protein, called GDNF, using a radical new drug-delivery method that sends it deep into the part of the brain where Parkinson's originates. There, a constant supply is administered by a pump implanted in the chest. So far, GDNF seems both to shield healthy brain cells from the disease and cause damaged cells to regenerate.

After just a few months of testing, says Kentucky investigator Greg Gerhardt, there is evidence of improvement in patients. In addition, British doctors reported last April that a similar trial in Bristol, England, improved muscle control of all five patients tested within a month of treatment.

These results, while preliminary, were welcome news for Amgen Inc. (AMGN ), the world's biggest biotech player and the only maker of GDNF. In the late 1990s, Amgen injected the protein into the brains of a number of Parkinson's patients. The drug showed no positive effect, and in 1999, Amgen managers dropped it from their development program. Doctors now believe the trial failed because GDNF never reached the region of the brain affected by Parkinson's. Amgen's enthusiasm cooled, but academic researchers at several institutions were convinced that GDNF held promise and kept working on their own. The University of Kentucky team, along with British neurosurgeon Dr. Stephen Gill, who led the Bristol study, figured out the new delivery method used in the current tests. Several other research centers are drawing up plans for similar trials.

The Kentucky scientists received a $5 million grant from the National Institute of Neurological Disorders & Stroke (NINDS), a division of the National Institutes of Health, to fund their trial. They also persuaded Amgen to collaborate. Today, Amgen CEO Kevin W. Sharer calls GDNF "one of the most promising products" in the company's pipeline, and Amgen is planning clinical trials of its own. Sharer cautions that GDNF is still in a very early testing phase: "We don't want to give false hope, but if this proves to be a success, it will be a wonderful advance."

Even a modest advance would be fantastic. Parkinson's is a progressive disease of unknown cause that affects as many as 1.2 million people in North America. The illness destroys the cells in a portion of the brain called the substantia nigra that is responsible for the production of dopamine, a chemical involved in muscle control. As the level of dopamine gradually drops, the disease's characteristic tremors start. These grow violent as the illness progresses and are accompanied by stiff limbs, frozen facial muscles, and eventually, rigid immobility. While generally not fatal, Parkinson's is gruesomely debilitating. Most victims are over 50 when the disease strikes, but about 10% are under 40; actor Michael J. Fox first experienced tremors when he was 29. There is no cure, only treatments to alleviate symptoms, which lose effectiveness over time and can have severe side effects.

Scientists have hoped that GDNF might provide relief ever since its function was discovered in 1993. GDNF stands for glial cell line-derived neurotrophic factor, a protein produced by the connective tissue, or glial cells, between the neurons of the central nervous system. Animal studies indicated early on that GDNF can not only halt but reverse the damage caused by Parkinson's. But GDNF is a large molecule that cannot be absorbed through the stomach or penetrate the protective blood/brain barrier, ruling out pills and shots. That seemed to leave only one option: delivering it directly to the brain, a risky venture in human patients.

In early Amgen trials, doctors injected GDNF into the lateral ventricle, a fluid-filled portion of the outer brain. The theory was that the fluid would carry the drug deep into the brain. Instead, says Gerhardt, director of the University of Kentucky's Parkinson's research center, GDNF stuck to the walls of the ventricle. Gerhardt and Donald M. Gash, who runs the university's magnetic-resonance-imaging center, spent five years trying to come up with a better way.

The result is a complicated delivery system built around a Medtronic Inc. (MDT ) implantable pump called the SynchroMed, in use since 1988 to administer chemotherapy and painkillers. For the Parkinson's trial, a small catheter is threaded into the substantia nigra using MRI coordinates to precisely map the route. The catheter is attached to a tube that runs under the skin, down the neck, to a SynchroMed implanted in the chest. About three inches in diameter, the pump contains four weeks' worth of the drug and is programmed to maintain a constant flow to the brain. It is refilled each month through a port attached to the pump.

Gerhardt says it's too early to draw conclusions since the first human patient began treatment only in May. However, in the October issue of the medical journal Brain, Gerhardt and Gash report that rhesus monkeys with Parkinson's showed a significant reduction in symptoms within three weeks of treatment--without any detectable side effects. Autopsies revealed a partial restoration of the dopamine-producing cells, indicating that brain cells could be prodded to repair themselves. An accompanying editorial called this "the first demonstration that GDNF infused directly into the brain" is effective in restoring dopamine function.

Still, many drugs over the years have proved effective in animals only to fail in humans. "GDNF is really promising," says Diane D. Murphy, program director of the neurodegeneration group at NINDS. But she is not convinced that infusion is the best approach. "You need constant, ongoing infusions, and that's a problem," she says. "We need a less invasive delivery method."

To that end, NINDS is funding a large, multicenter study that will investigate the use of gene therapy as a way of getting GDNF into the brain. Jeffrey Kordower, research director at the Research Center for Brain Repair in Chicago, was one of the pioneers of this method. He uses an inactivated virus to carry a gene into cells of the brain, which it prods into producing more GDNF. In Kordower's studies, monkeys with Parkinson's that underwent gene therapy performed at near-normal levels a week after treatment. The genes remained active up to eight months after injection, and again, dopamine-producing cells were restored.

Researchers are also trying to develop stem cells that would turn into GDNF-producing cells when placed in the brain. But both gene therapy and stem cells present serious safety issues--recent human trials of both methods in other diseases have produced disastrous side effects. Consequently, human trials are still years away. For now, the direct-infusion method is the most promising. If it works, says Gerhardt, the same procedure could potentially be adapted to treat a variety of brain diseases. Now that scientists understand that brain cells can regenerate, there may finally be a glimmer of hope in treating the horrible ailments that destroy them.

By Catherine Arnst in New York, with Arlene Weintraub in Los Angeles

— With assistance by Arlene Weintraub

    Before it's here, it's on the Bloomberg Terminal.