Some patients who underwent the first full trial of the surgical technique have developed alarming side effects which doctors say they cannot reverse.
While the treatment appears to have helped a small number of patients, it was of no benefit at all for patients over the age of 60. And in 15% of cases, the patients ended up with worse symptoms than they had before they took part in the trial.
For the first year after undergoing surgery, the patients appeared to be progressing well. However, many patients then began to develop distressing symptoms, including an uncontrollable jerking of the head, writhing and throwing of the arms.
The researchers attempted to treat Parkinson's by replacing the brain cells whose death caused the original symptoms with immature cells taken from aborted foetuses.
Excessive levels
The cells were successfully implanted into the brain of sufferers, but once there continued to multiply.
This led to over-production of the brain chemical dopamine, which plays a vital role in co-ordinating movement.
Levels of dopamine are depleted in Parkinson's patients, causing the problems with movement associated with the condition. However, excessive levels of the chemical can also produce violent, uncontrolled movement.
The failure of the trial casts new doubt on whether it is safe to use immature cells to try to repair damage to parts of the body - an area which is currently the subject of much research.
Research 'must continue'
Dr Paul Greene, a neurologist from the Columbia University College of Physicians and Surgeons in New York, was one of the researchers who conducted the trial.
He said five patients were now unable to control their movements. He told the New York Times newspaper: "They chew constantly, their fingers go up and down, their wrists flex and distend. It was tragic, catastrophic. And we can't selectively turn it off."
One man now has to be fed through a tube because he can no longer eat.
However, Dr Gerald Fischbach, director of the Neurological Disorders and Stroke which funded the study, insisted that research in the field should continue.
He said: "Clinical research of this sort is difficult and it is risky. I don't think because of unanticipated adverse effects that we can afford to call a halt to it.
"I think the patients themselves would feel deprived of the fruits of a lot of fundamental research."
Lessons learned
Robert Meadowcroft, director of policy and research at the Parkinson's Disease Society in the UK, told the BBC: "What appears to have happened is that the treatment has produced even worse symptoms than the patients started with.
"It is very, very distressing for those people who clearly have taken a great risk in undergoing this clinical trial and it has had devastating results for them."
Mr Meadowcroft said the results of the trial were a "major setback" in the search for a cure for Parkinson's.
But he added: "We must not panic, we must continue with laboratory-based research, and when we move into clinical trials on patients in this country - perhaps in three years - I hope we will be able to learn from these problems in the US."
The researchers used cells from four foetuses, giving some of the patients "sham surgery" where holes were drilled in their skull but no cells implanted, to create a comparison group among the 40 people, aged 34 to 75.
The research is reported in the New England Journal of Medicine.