Steven Goldman and his colleagues experimented with rats which had a condition mimicking Parkinson's disease. When treated with mature neurons which had been created from human embryonic stem cells, they recovered all their lost motor function. So far, so good.

However, ten weeks after the transplant, the rats' brains were examined. To the dismay of the researchers, only about one-fifth of the transplanted cells were now producing dopamine (the chemical lacking in patients with Parkinson's). The rest had begun dividing uncontrollably into benign brain tumours.

Although a stem cell scientist quoted in the journal Science, Evan Snyder, of Burnham Institute of Medical Research, declared that further fine-tuning of the research will eventually find "the recipe to make it perfect", the language of the article was more pessimistic. "These promising results were unfortunately accompanied by several potentially disturbing outcomes," it said. First, the number of neurons producing dopamine fell sharply after one month, even though they had been "highly enriched" and differentiated cells. Second, "their persistent, uncontrolled and grossly homogeneous expansion over a 10-week span before the animals were killed nonetheless suggested graft-associated tumorigenesis". And it seemed clear, too, that the problem was not contamination by undifferentiated embryonic stem cells. The very cells which were supposed to cure caused the tumours.

What's going on? The first explanation is that nothing is. Dr James Wells, of Long Beach, California, told the magazine that most of the women who committed suicide all had their implants decades ago. "It's a very different world now," he says. "The implants are better, how we assess the patients is better, and implant failure rate is lower." However, epidemiologists are not convinced. A remote possibility is that leaks from the implants trigger changes in brain chemistry. Another is that women with breast implants are more likely to be substance abusers.

But the most plausible explanation is that the women who seek cosmetic surgery come with personality traits or psychiatric disorders which their doctors fail to detect. Indeed, a Danish study found that 8% of women with breast implants had earlier been in a psychiatric hospital. As well, an estimated 6% to 15% of patients who seek cosmetic surgery are believed to have a condition called body dysmorphic disorder, more commonly known as "imagined ugly syndrome". These people are obsessive about non-existent flaws in their personal appearance. They are 45 times as likely to commit suicide as normal people -- more than twice as much as people with major depression.

A British specialist disputed the results. "The problem I have is not what happens to mothers at 50 to 55. What worries me is what happens when their children are 18 and they are in their 70s and 80s," said Bill Ledger, professor of obstetrics at the University of Sheffield. "I am not in favour of banning things, but I am concerned for the welfare of a child whose parents are as old as its peers grandparents."

Nearly all of the funds for the Coalition for Lifesaving Cures have come from a single philanthropic couple, Jim and Virginia Stowers, who have bankrolled the Stowers Institute, a first-class new medical research facility in St Louis. Support is broad for the initiative, claim supporters, even though the Stowers have given 97% of the funding so far. But opponents complain that it is not a broad-based citizen's initiative. "This isn't grass roots," says a political scientist at the University of Missouri at Columbia. "This is Astroturf."

So an interview with Caplan in the latest Technology Review, a magazine published at the Massachusetts Institute of Technology, helps to understand the forces shaping bioethical debates. "I'm a consequentialist," Caplan tells interviewer David Ewing Duncan. "I'm looking at outcomes. I'm trying to decide if a particular policy -- such as allowing surgeons to do face transplants -- would do more harm than good."

Fellow consequentialists include Princeton philosopher Peter Singer, but Caplan regards him as too rigorous in his logic. "He says that if animals are smarter than retarded children, then experiment on retarded children. I'm not willing to trust any theory that far. In general, I'm not looking for fundamental truths when I discuss ethics. What matters is what is most practical at a given time. I ask, 'What are the benefits and costs?' And I understand that the answer will change over time."

The social implications could be immense, especially for infertile couples. Artificially produced eggs would mean more choice, says the New Scientist. "A limitless supply of eggs could be used to produce hundreds of embryos for any one couple, making it feasible for would-be parents to use genetic screening to choose several desirable traits as well as ensuring embryos are free from disease- causing mutations." It would also be possible for gay and lesbian couples to have their own genetically related children (although lesbians could not have boys.)

Progress is slow. So far, few artificial eggs have developed far after fertilisation. "Oocytes are very sensitive cells and it doesn't take much to send them off the rails," says Alan Trounson, of Monash University. "Getting them to mature and undergo meiosis is a big hurdle." Mice have been born from artificial sperm, but these were poor swimmers and had to be injected into the eggs. All of the resulting mice died within five months and all of them had serious health problems.