In a study published by Nature Medicine, the University of Amsterdam team report on their detective work, launched after a seven-month-old girl was admitted to their hospital in April with bronchiolitis.

    Tests for the common cold, flu and other well-known viruses all proved negative.

    "At first, we were worried that the child may have caught the virus from an animal because she had been to the zoo the previous weekend," said the research supervisor, human retrovirology expert Ben Berkhout. Those fears proved unfounded, for the team then "did a little amplification trick" of the viral sample and "ended up with a couple of fragments that looked like coronaviruses but not one of the known ones," he said.

    Closer examination of its genetic code showed it shares about two-thirds of its nucleotides with the three established coronavirus types-in other words, it was an independent member of the coronavirus family that had never been spotted before.

    The team then tested stored samples taken from other admissions and found that seven other patients with respiratory problems-seven percent of admissions-had been infected by the same virus.

    Four of the seven were children under one year old and three were adults. Two of the grownups had a damaged immune system-one had had a bone-marrow transplant and the other had AIDS.

    The findings suggest "the virus is already widely spread within the population," the research, led by Lia van der Hoek, says.

    "This virus probably has spread worldwide, I would say, and it's causing perhaps the symptoms of the common cold but in the younger ones, it's causing more severe respiratory problems," Berkhout said. "About 20 to 50 percent of respiratory problems do not have a viral cause, so we now can fill in a significant fraction of those unknowns."

    Berkhout admitted it was "surprising" that this virus had never been found before, but part of the problem was the lack of tools to detect new agents easily. "It's not so easy to find a new virus when you suspect there should be something but you don't know where to start. You need new technology to find new viruses and that's what we had in this study," he said.

    The other coronavirus groups are HCoV-229E and HCoV-OC43, first identified in the 1960s.

    At least 10 candidate vaccines are under development in Asia, Europe, and the United States, according to Marie-Paule Kieny, director of the Initiative for Vaccine Research at the World Health Organization (WHO). Even so, "it will take a few years" before any of these are proven safe and effective for the public, she said.

    Heading the pack is a fast-tracked prototype from China that uses a killed virus to provoke an immune response. The trial vaccine has been successfully tested on monkeys and was ready to be assessed for safety on 30 human volunteers.

    This is Phase I of the human testing procedure. Phases II and III involve much larger groups to test again for safety and efficacy.

    "Making a vaccine that's safe and can be put into people is usually something that takes a minimum of five years," said Maria Zambon, head of the flu laboratory at Britain's Health Protection Agency.

    The time can be shortened if there is a worldwide peril.

    But express-testing new drugs can also come at the cost of safety, as experiences showed with the very first anti-HIV virals that for many patients had grim side effects.

    The bad news for SARS: The scare factor that drives vaccine funding and public health policy has receded after SARS was successfully contained.

    Zambon said containment of the outbreak meant corporate interest in a SARS vaccine-a long and costly quest with an uncertain reward-has "definitely fallen back."

    Despite this, funding lines have been opened up by the US National Institutes of Health (NIH) and the European Union's Executive Commission, dangling money for joint public-private sector initiatives.

     "I would expect to see in the next year or two a number of different academic sectors/companies reporting their progress, be that in animal vaccine development or vaccine for SARS in humans," Zambon said.

    The "killed virus" approach adopted by China is a classic, swift design that hopes to induce a broad spectrum of antibodies and be long-lasting, said Sylvie van der Werf, a professor at the Pasteur Institute in Paris.

    The downside, though, is that sometimes a killed virus vaccine can act on the immune system in unexpected ways. "This is what we saw in a vaccine for feline infectious peritonitis. It encouraged infection rather than protection," she said.

    There remain many unknowns about the coronavirus that cause the disease.

    Among them are the animal "reservoir" where the virus naturally holes up; how the SARS virus works once it enters the body; if, like the flu virus, the SARS agent mutates steadily or suddenly; and why some individuals are mysteriously at greater risk to it than others.

    Understanding that will help the lab engineers seeking a formula that will prime antibodies and killer cells against the virus. It will also help decision-makers who must determine whether vaccination efforts should be better directed at animals rather than humans, or at individuals who are particularly vulnerable to the virus or most at risk of transmitting it, Zambon said. AFP

    The samples came firstly from blood donated by 400 healthy people between March and May last year; from 395 hospital patients, adults and children, who did not have pneumonia, which is the worst symptom of SARS; and finally from 33 health workers at the height of last year's SARS epidemic in Hong Kong who did not have any SARS symptoms at all. Around four percent of these samples showed antibodies for SARS, indicating that that person had, at some point, been infected by the virus.