Twenty four people accidentally given blood transfusions contaminated with the human form of mad cow disease are now at "substantial" risk, an expert said today.

Professor John Collinge made the assessment after investigating the third case of a person known to have acquired variant Creutzfeldt-Jakob Disease (vCJD) from a blood transfusion.

Writing in The Lancet medical journal, he said blood transfusions seemed to be an efficient route for transmitting the infectious prion proteins believed to cause vCJD.

The first two cases of vCJD infection linked to transfusions were reported in 2004.

Both patients had received blood from donors who later developed the disease. One recipient developed signs of vCJD after six years, and died 13 months later.

The second did not develop symptoms and died from an unrelated cause five years after the transfusion. A post mortem then revealed evidence of vCJD.

The third patient, a young man of 23, was the first to be diagnosed with the brain disease while still alive.

Like the others, he was given blood from a donor who later developed vCJD. Seven-and-a-half years after his transfusion he was referred to the NHS National Prion Clinic at the National Hospital for Neurology and Neurosurgery in London. There, his symptoms were confirmed to be caused by vCJD.

The patient joined a drug trial testing a potential treatment for the disease in 2004, but died a year later aged 32.

Initially 66 transfusion patients were known to have received infected blood from donors who developed vCJD.

Many were seriously ill from other causes and subsequently died for reasons unrelated to the disease.

Today a total of 24 remain alive, but facing an uncertain future. VCJD has a long incubation period which may vary according to an individual's genetic make-up.

Prof Collinge, from the Medical Research Council's Prion Unit, said today: "That three individuals from this small group of people that we know to have been exposed through blood transfusion have already developed vCJD infection suggests that the infection may be efficiently passed by this route, so the risk to remaining individuals is likely to be substantial.

"A national tonsil tissue screening study being performed by the Health Protection Agency may soon give estimates of the number of people who are silently infected with prions. This information is vital for public health planning given the relative ease with which prions seem to be passed on by blood transfusion."

He said it might be possible for a person to be silently infected with vCJD for more than 50 years before developing symptoms.

During this time a "carrier" posed a potential risk to others, by transmitting the infection through donated blood or contaminating surgical and medical instruments.

The incubation period was likely to be much less when vCJD was passed from one person to another in blood, rather than being acquired by eating contaminated beef.

Without having to jump a "species barrier" from cows to humans, vCJD spread via a blood transfusion could develop in just six or seven years, said Prof Collinge.

Tests on the latest victim revealed that his tonsils were infected with vCJD prions.

Previously it had not been certain that looking at tonsils could identify patients infected by a blood transfusion.

Prof Collinge said: "Although we do not yet have an effective treatment for any form of CJD, a reliable tonsil test could allow people with vCJD to access experimental treatments early."