News

The progress of science is paved with stories of high hopes and heartbreaks. But in a busy lab at the University of Rochester the two extremes have met in one dazzling yet devastating experiment.

Researchers there have for the first time essentially cured rats of a Parkinson's-like disease using human embryonic stem cells. But 10 weeks into the trial, they discovered brain tumours had begun to grow in every animal treated.

"Here we have this method that works so well to reverse the symptoms of Parkinson's," said lead investigator Steven Goldman, "But no matter how you look at it, it's an expanding mass and that's bad news."

None of the cells growing out of control were cancerous tumours. But as Dr. Goldman pointed out, "In the brain, nothing's benign."

The work, published today in the journal of Nature Medicine, is a sobering setback for plans to use stem cells from human embryos to grow tissues for human transplant.

"My hopes are still high, but I think this injects real caution," said Dr. Goldman, who spent four years on the experiment and a 23-year career building up to it. "Some folks are portraying this as imminently useful and it's not. There's still a lot that has to be sorted out."

By definition, human embryonic stem cells have the almost mythical, immortal power to grow and divide indefinitely as they become the various tissues that make up the body. As a result, scientists have always known that any stem cell therapy could result in an uncontrolled growth of cells that could give rise to cancer.

But that risk has remained largely theoretical since there have been few attempts to transplant tissues grown from stem cells into live animals. The work is difficult, expensive and tricky to pull off when several countries -- Canada included -- have enacted tough laws that limit research on stem cells from human embryos.

"A lot of the representations of stem-cell research have resulted from the initial excitement and speculation of what can be achieved. But we're still in that early stage, we haven't seen real clinical breakthroughs," said Tim Caulfield, director of the Health Law Institute at the University of Alberta. "This [experiment] shows the incredible potential of the field, but it also sheds a more realistic light on the near-future potential."

For Mick Bhatia, scientific director of the Cancer and Stem Cell Research Institute at McMaster University, it's a bit of déjà vu. In 2004, he succeeded in growing human blood cells from embryonic stem cells but found transplanting them into mice wasn't simple.

"I pushed the program back," Dr. Bhatia said. "We need to do more on the basic biology."

Still, few scientific fields are hotter than stem cells as researchers everywhere investigate the possibility of using them to grow replacement parts -- cardiac cells for heart patients or islet cells for diabetics.

Parkinson's has been considered one of the prime candidates for a stem-cell therapy because just a single cell type is needed -- one that produces dopamine.

Neurons that make dopamine are crucial for movement and degenerate in people with Parkinson's, often leaving them stiff, unable to control their physical gestures and suffering from tremors.

Twenty years worth of studies have tried treating Parkinson's with dopamine-cell transplants, in rodents, primates and people. In the 1990s, after a Swedish study found some benefit to patients who received dopamine-cell transplants from aborted fetal tissue, large human trials began in both Canada and the United States.

"These failed," Dr. Goldman said. "It made things worse; patients suffered movement abnormalities." In part, he explained, this was because the transplants contained all sorts of cells. Less than 10 per cent of the cells, and in some cases, less than 1 per cent, produced dopamine.

For this reason, Dr. Goldman, a neurologist and chief of cell and gene therapy at the University of Rochester, and many others, considered embryonic stem cells a possible source of growing only the cells that would be needed and they were right. But no one had been able to grow enough for a transplant.

In this experiment, however, Dr. Goldman and his team overcame the volume problem by tricking the embryonic stem cells into behaving like they were growing in the developing brain.

To do this, they harvested glial, or brain-support cells, from the precise brain region of an aborted fetus that would, at 11 to 22 weeks gestation, trigger the development of the dopamine neurons needed.

The researchers then used a retrovirus as a courier to deliver into the DNA of these cells a gene, known as telomerase, which would immortalize them. This way, Dr. Goldman explained, the support cells would continue to grow endlessly and continuously give off the chemical cues to keep stem cells maturing into dopamine cells.

The dopamine cells had first been grown from the limited number of human embryonic stem-cell lines that U.S. President George W. Bush made available to government-funded researchers in 2000.

Culturing the immortalized glial cells alongside the stem-cell derived dopamine neurons -- although not touching each other -- resulted in a growth of dopamine cells three to 10 times what is normally seen.

But Dr. Bhatia, who read the report, said this step might have contributed to the uncontrolled growth of the cells.

Even though the immortalized glial cells and dopamine cells were not touching each other, there could have been chemical cues exchanged that affected the implanted cells, he said.

"The number of cells they're getting is incredible," said Dr. Bhatia after reading the report. "But at what cost did they gain that efficiency?" In a series of trials, the researchers next implanted tissues of 500,000 cells each into the midbrain region of dozens of rats, which, as a result of a chemical injection, suffered a condition similar to Parkinson's. Seventy per cent of the cells transplanted made dopamine.

Before the treatment, rats with the Parkinson's-like disease suffered from constant tremors and a lack of co-ordination. But four weeks after the transplants, Dr. Goldman said, they showed a marked improvement from the control group of untreated rats.

"It was really a complete recovery on the part of the animal. By six to eight weeks, they're normal," he said. "That's a more powerful effect than ever seen before."

But at 10 weeks, when the animals were autopsied, the researchers discovered that the implanted tissue had given rise to more than just dopamine cells.

Several cells had begun to divide at a fairly steady pace, the hallmark of cells growing into a tumour.

Where the dopamine cells once made up 70 per cent of the tissue implanted, at the autopsy they made up only about 25 per cent.

"They were undifferentiated neural cells that were expanding and dividing, and those are cells you don't want there. You don't need to be a neuro-oncologist to say, that's the start of a tumour," Dr. Goldman said.

Dr. Bhatia also raised the possibility that the years-old and scant stem lines available to government researchers in the United States may also have had tumourigenic properties from the start that skewed the experiment.

Dr. Goldman and his team are now redoing the experiment on the basis that neural cells other than dopamine-cells in the transplanted tissue led to the tumour growth.

"We are going to have to absolutely purify the cell type of interest," Dr. Goldman said. "This really pushes this [kind of transplant work] back in terms of clinical use.

"It's not ready for prime time, that's for sure."