Aids research fault is not likely to be accident, say experts

Experts have cast serious doubt on suggestions that pure accident led to the faulty research results which prompted a team of Manchester scientists to claim, wrongly, that the world's first Aids patient died in 1959.

Following an investigation by The Independent, which first revealed doubts about the research, the experts - who include an Oxford professor - say there is only a 3 per cent chance that faulty tests which led to the claims arose through bad luck.

In 1990 a team of scientists at Manchester University found HIV, the virus that causes Aids, during retrospective analysis of tissue samples from David Carr, a 25-year-old apprentice printer who died of a mysterious illness in 1959. None were found in "control" samples from another person who died at the same time.

This led them to make the world-famous claim in the medical journal the Lancet that July that Mr Carr was the "world's first confirmed case of Aids" Based on this, a number of scientists theorised that Aids was an ancient disease, rather than a new one which had arisen in the past few decades.

But the Manchester claim was subsequently exploded by The Independent, which revealed in March 1995 that American researchers had determined that the strain of HIV found in the tissues evolved too recently to have existed in 1959. In January, Andrew Bailey and Gerald Corbitt, two of the three scientists who submitted the paper, wrote to the Lancet to admit "we can find no evidence ... to suggest that the 1959 Manchester patient carried [HIV]".

Edward Hooper, a medical researcher, and Professor William Hamilton, a geneticist at Oxford University, suggest two possible causes. The error which led to HIV being found only in Mr Carr's samples could either have occurred before the samples were sent to the laboratory where they were ana- lysed, or by a mistake in breaking the "double blind" coding (which prevented them knowing whether the sample came from Mr Carr or from a "control").

Few alternative scenarios appear feasible. Contamination of samples from Mr Carr by an HIV-infected knife would only be explicable if "the knife supposedly wiped itself clean" on the control samples, they say.

But Professor Hamilton said: "The fact that there are at least three human genotypes involved suggests it cannot have been a simple contamination."

The third author, Dr George Williams, the Manchester University pathologist who conducted the post-mortem examination on Mr Carr, has repeatedly denied this. "I'm utterly, absolutely confident of the authenticity of that material," he told The Independent last year.