When British scientists announced two years ago that they had developed a drug that could potentially cure obesity, it created excitement around the world.

When British scientists announced two years ago that they had developed a drug that could potentially cure obesity, it created excitement around the world.

Researchers raced to check claims that a drug based on a hormone in the body could promote weight loss by cutting food intake by a third.

Headlines spoke of a cure for the "third helping" after the head of the laboratory that made the discovery, Professor Stephen Bloom of Imperial College London and the Hammersmith Hospital described how the drug - PYY3-36 - curbed hunger by telling the appetite centre in the brain when the stomach was full.

At least two drug companies took up the chase and millions of dollars have been spent in experiments on the drug.

But the "breakthrough" may turn out to be an apparition. Professor Bloom's research has been questioned by more than 40 scientists from 15 international research centres, which report that they have been unable to replicate his findings.

So frustrated have they become that they have taken the unusual step of writing to the journal Nature, which published the original claims, to express their disappointment.

The dispute revolves around the race to create an obesity wonder drug, one of the most important - and potentially lucrative - quests in medicine.

Professor Bloom and his colleagues have spent more than 20 years studying the causes of obesity and the hormones that control appetite.

When his team published claims in 2002 that the hormone decreased food intake and caused weight loss in mice and rats, Professor Bloom announced that they had found the "hormone for satiety". An infusion of the hormone given to human volunteers curbed their appetite for 12 hours so that, when tempted with a buffet offering chicken curry, chocolate bars and other delights, they consumed a third less calories than volunteers given a placebo infusion.

Professor Bloom said more than a billion people around the world were extremely overweight and a solution to the problem was vital. "It may be possible," he said, "to identify foods which cause the release of more PYY3-36, helping to naturally limit appetite, or it may be possible to create a tablet with similar effect providing an excellent, natural and safe, long-term treatment for obesity."

But then the problems began. A group of 42 scientists from leading obesity-research institutions and universities across Europe and America, tried to the replicate the findings in rodents.

Now, they say their failure to do so "calls into question" the potential value of an anti-obesity approach based on the administration of PYY3-36.

Led by Matthias Tschöp of the University of Cincinnati, the group says that the original aim of repeating Professor Bloom's experiments was to accelerate the development of a new treatment for obesity.

Like Professor Bloom, the group injected PYY3-36 into the bodies of young rats. But they found the rats ate no less than a control group injected with salt solution. Yet Professor Bloom's group had reported a decrease in weight gain of some 20 per cent.

When the announcement about PYY3-36 was made two years ago, it was potentially the most important breakthrough in treating obesity because the drug promised to be at the centre of the brain's ability to control appetite in both slim and overweight people.

"Before PYY3-36, we thought there could be no single treatment for obesity. Then, for a moment, it seemed like we had a magic bullet for all patients," said Dr Tschöp.

When he found that other researchers were having similar difficulty replicating Imperial College's results, they decided to collaborate. They bought rats from the same source as Professor Bloom, bought food from the same supplier, and injected the animals at the same time of day with the same kind of syringes filled with PYY3-36 made by the same chemical supplier.

"The results we generated were entirely unexpected. We have been unable to replicate their results," Dr Tschöp said.

"Although the reasons for this discrepancy remained undetermined, an effective anti-obesity drug ultimately must produce its effects across a range of situations."

Professor Bloom insisted the doubters were wrong. He said his experiments had been replicated many times by his own group and even by two outside drug companies. He pointed out that his research had come under attack in the past and had later been proved correct.

"Our results are completely convincing. It's not a problem replicating the work. It's a question of doing it properly," Professor Bloom said.

Handling the laboratory rats and mice during the experiments can cause stress which can put the animals off their food and so interfere with the action of PYY3-36, he said. The other scientists had not taken this into account, he suggested.