AIDS researchers and charities welcomed the announcement on Tuesday of a "drugs cocktail" that offered a significant breakthrough in treatment for the deadly disease.

Britain's Medical Research Council announced that combining the common AIDS drug AZT (zidovudine or retrovir) with either one of two related drugs - ddI or ddC - prolonged the lives of up to 38 percent of HIV sufferers over two years."This is very good news for people with HIV and AIDS," said Nick Partridge, spokesman for Britain's AIDS charity the Terence Higgins Trust. "We are still a long way from a cure, but these results renew hope."

The results were so overwhelming the researchers rushed to publish their findings while in the middle of the study.

"Our independent ethics board advised us to publicize the results," said Dr Tim Peto, a specialist in infectious diseases at John Radcliffe Hospital in Oxford, England, and one of the coordinators of the study.

Another coordinator, Dr Brian Gazzard of London's Chelsea and Westminister Hospital, said HIV patients starting drug treatment should take the combination straight away.

"We felt that in general doctors should know and the public should know this is true," he said. "I think what's clear is that people now starting therapy should start on combinations."

The so-called Delta trial, which started in 1992, involved more than 3,000 people with HIV in Britain, Ireland, Australia, France, Germany, Italy, the Netherlands and Switzerland.

They were randomly assigned one of three different treatments - AZT alone, AZT with didanosine (ddI) or AZT with zalcitabine (ddC) - all similar drugs that attack the AIDS virus's ability to work and replicate itself.

Of the patients who had never taken AZT before, 17 percent who took AZT alone died, compared with 10 percent who took AZT with ddI and 12 percent who took AZT with ddC.

A second group of patients had been taking AZT already. In this group, death rates did not differ significantly.

"The combined results yield a reduction in mortality of about 25 percent in favour of combination therapy," the research council said, adding that the reason the second group did not respond as well was not clear.

Peto said the researchers would be looking urgently at why this should be.