IT WAS the drug that held out hope to people carrying the world's most feared virus. It had the power to move share prices by millions. What it could not do was help people facing AIDS.

This weekend the truth about AZT is in the open: a comprehensive trial, so big it equals all the other research put together, shows that the drug which dominates AIDS treatment has no effect in delaying the onset of the disease. After all the promise and the profits, AZT has nothing to offer people with HIV.

The findings came in the final report on the Anglo-French Concorde trial, published yesterday in The Lancet. Some 1,749 patients with HIV, but who showed no symptoms, were given either the drug or a placebo. There was no statistical difference in the progress of the two groups: after three years 18% had AIDS or were dead.

The results leave a terrible void for the 12m people worldwide said to be infected with the virus, and crush any remaining hopes that AZT might delay the onset of symptoms. They also raise questions as to how those hopes were fuelled in the first place.

Doubts about AZT were first revealed by The Sunday Times five years ago. A painstaking investigation showed that AZT had been rushed to market on the back of a flawed study that was supposed to demonstrate its effectiveness.

The American Food and Drug Administration (FDA), responsible for protecting the public from risk, had been aware of flaws in the trial, but gave AZT approval. Documents obtained under the American Freedom of Information Act showed that records compiled during the trial had been altered, giving the drug a more favourable record; "multiple deviations" from the terms of the study had occurred; and FDA investigators had argued for data from one centre to be dropped entirely from the results. A senior FDA official believed AZT should not be granted a licence, but was overruled.

The doubts did nothing to inhibit Wellcome, AZT's maker, from promoting its drug. Patients with HIV, but without AIDS symptoms, were the new target. They are worth more money because there are more of them and because they have longer to live.

To show the drug's usefulness to this lucrative group, Wellcome trumpeted a big American trial called Protocol 019. The trial was halted in August 1989, after less than two years, on the grounds that it had already shown such benefit to HIV-positive people it would be unethical not to give the drug to all who wanted it.

Such "benefit" was judged only by time free from disease. A new analysis of the trial data, however, reaches a similar conclusion to Concorde: that AZT is essentially useless.

The original results were announced with a fanfare by the National Institute of Allergy and Infectious Diseases, which sponsored it with Wellcome's support. In London, The Independent newspaper gave its front page to the findings, under the headline "AIDS drug offers lease of life".

The very different picture painted by last month's analysis, in the New England Journal of Medicine, comes after investigators paid more attention to the drug's side-effects. These can include anaemia, liver damage, fatigue, nausea, headaches and sometimes a collapse in white blood cells, making patients more prone to disease.

The researchers looked at the average time patients experienced neither a progression of disease nor an adverse effect. Those treated with low doses of AZT were found to suffer a reduction in quality of life "due to severe side-effects of therapy" that approximately equalled any benefit from slowing down the disease; people on higher doses suffered even greater side-effects, outweighing the supposed benefit.

Dr Peter Duesberg, the American virus expert who has claimed for years that AZT is not a rational therapy, says it is clear that the original claims were completely ill-founded. "The opposite interpretations of the same data lead me to conclude that those responsible are not acting as scientists; they are acting as politicians.

"When the time is ripe to say that AZT is detrimental, that it actually hurts, the interpretation will change again."

For patients with AIDS-related symptoms, AZT will continue to be prescribed: the consensus remains that it gives a temporary benefit.

For those without symptoms, hope centres on combinations of drugs, or on other approaches such as gene therapy. However, Professor Ian Weller, of the Middlesex hospital in London, who was the principal British investigator in the Concorde trial, is alarmed by the drive to give AIDS patients an AZT drug cocktail as if it were already an established therapy.

"There's a suspicion of more toxicity if you combine it with other treatment, and we are a long way from showing an important clinical benefit, or that it is safer than AZT on its own," he said. "There are physicians who are jumping the gun."

As late as Thursday, Wellcome was insisting that AZT "remains the best weapon we have to slow the progress of the disease". Dr Trevor Jones, its research director, said: "The question is where in the course of the disease you begin." *