Scientists have found a way to remove disease-causing proteins from infected animal blood, which they hope may fight the human form of mad cow disease.
The filtering technique could be used to treat human blood
The UK and US researchers identified a molecule which removed the prion proteins from blood infected with scrapie, the Lancet reports.
Scrapie, which can affect sheep, is related to variant CJD. It is hoped the technique could also treat human blood.
But a blood transfusion expert said it first had to be properly assessed.
Prion diseases, including bovine spongiform encephalopathy (BSE) - or mad cow disease - and variant Creutzfeldt-Jakob Disease (vCJD), are fatal neurodegenerative diseases that occur after years of incubation with no apparent symptoms.
In the UK, three people have died from vCJD after receiving blood from donors who later went on to develop the disease.
Globally, there have been 200 cases of vCJD reported, 164 of which were in the UK.
'Removal is best'
The researchers, including scientists from Cambridge University, suggest removing the infectious ability of prions - mis-formed proteins - might be one of the best ways of reducing the risk of vCJD transmission through blood.
They screened millions of molecules and found that one, called L13, binds prion protein (PrP), removing it from the blood.
The team checked this by passing 500 millilitres of scrapie-infected hamster blood through a filter that removes white blood cells - something which is done to all blood donated from transfusion in the UK.
When they injected the blood into 99 hamsters, 15 became infected with scrapie.
But when they passed treated blood through devices containing the removal molecule and injected 96 hamsters with it, they found that none became infected.
L13 was also found to bind to PrP from human infections of vCJD, suggesting that it may also remove prion infectivity from human blood.
Dr Robert Rohwer, from the Veterans Affairs Medical Center, at the University of Maryland, Baltimore, who was involved in the study, said: "Removal of vCJD infectivity by adsorption gets around the extremely difficult problem of detecting the very low concentrations of these agents in blood, especially during the long asymptomatic period when people donate.
"For these disease agents, removal may be our best or only option."
Writing in the Lancet, Marc Turner of the Edinburgh Blood Transfusion Centre, said there were concerns about how best to check the technology worked, as there is nothing to compare them against.
There are also worries that passing blood through filters may cause blood to be caught up in the device and therefore lost, or could alter the blood's properties.
He added: "The UK and Irish Blood Services have produced quality, efficacy and operational specifications for such filters, and are considering an assessment programme that will include independent investigation of efficacy and clinical safety studies.
"Until these technologies can be clinically and operationally assessed, the best protection against the uncertain risk of transfusion-associated prion disease remains in ensuring that blood products are used only if needed - and that the uncertainties surrounding potential risks are communicated effectively to patients and to the public at large."