Sheep threat fenced out Estimates of CJD risk from sheep remain woolly.

10 January 2002 TOM CLARKE

If mad cow disease has spread to British sheep, the ongoing risk of human infection from eating sheep products may be greater than the existing threat from cattle, predicts new research.

The risk of getting variant Creutzfeldt-Jakob disease (vCJD) - the human form of bovine spongiform encephalopathy (BSE, or mad cow disease) - from sheep could be negligible. Or it could boost the total possible number of deaths (including those caused by cattle) from vCJD to 150,000, researchers calculate1.

The huge variability in the possible number of deaths reflects how little concrete data there is on BSE in sheep and its transmission to humans. As of December 2001 there have been 102 confirmed deaths from vCJD in Britain.

A comprehensive, five-year experiment to find evidence of BSE in British sheep ended in spectacular failure last year when sheep brain samples were found to be contaminated with cow tissue2. Assuming that BSE had first jumped from cows to sheep, epidemiologist Neil Ferguson and colleagues at Imperial College in London attempted to estimate the likely risk of human vCJD infection.

Using data for animals that were experimentally infected with BSE, they created a computer model to calculate the best- and worst-case scenarios for the spread of BSE within the sheep flock and then on to humans. "We cannot exclude the possibility that the epidemic is very large," says Ferguson. He adds, however, that the worst-case situation is by far the most unlikely.

First, it is possible that BSE never infected sheep. "There's nothing to indicate that there's something out there," argues Rowland Kao, an epidemiologist at the University of Oxford.

Records of what sheep were fed at the height of the BSE epidemic show that they were probably rarely exposed to BSE-infected cow material. "We know that at least one million cattle were infected with BSE; in sheep it is about 100 times less than that," estimates Ferguson.

The worst-case scenario also assumes that BSE can spread from sheep to sheep in the same way as scrapie, a similar disease that occurs naturally in the flock. BSE cannot, however, pass naturally from cow to cow.

It is also virtually impossible to calculate accurately the possible number of deaths from vCJD. Little is known about how long it takes for vCJD to kill people, or whether the disease can cross the species barrier from sheep to humans. "The numbers themselves are not that important," argues Kao.

Ferguson claims that his calculations put absolute limits on the likelihood of infection, and point to where more data is needed. "It highlights the uncertainties," he says.

Having prepared techniques for calculating risk, it will be possible to assess the danger posed by sheep as soon as data become available on the prevalence of BSE in the flock, argues Ferguson. "We now have a suite of analytical tools to analyse any data coming through," he says. Such data could be available within a year.

References Ferguson, N. M., Ghani, A. C., Donnelly, C. A., Hagenaars, T. J. & Anderson, R. M. Estimating the human health risk from possible BSE infection of the British sheep flock. Nature, 413, 709, (2002).
Butler, D. Brain mix-up leaves BSE research in turmoil. Nature, 413, 760, (2001).