A second baby has been cleared of HIV, almost a year after a similar case with a baby in Mississippi was announced. The second baby, who was treated in California, is still being given anti-HIV drugs, but blood tests can no longer identify virus particles capable of replicating. Both infants were given antiretroviral drugs almost immediately after birth.

Work in monkeys, also presented at the Conference on Retroviruses and Opportunistic Infections in Boston, suggests that the window for successful early intervention is narrow – a delay of just three days made a huge difference in the way the monkeys responded to treatment.

Usually, newborn babies with suspected HIV infections from their mothers are not treated until weeks after birth when infection has been confirmed. Audra Deveikis of the Miller Children's Hospital in Long Beach, California, gave the baby antiretroviral therapy four hours after birth because her mother, who has advanced AIDS, had not taken the antiretroviral drugs she had been given.

Later tests on spinal fluid confirmed that the California baby – now 9 months old – had indeed been infected.

Speedy treatment

This isn't the first time early intervention has paid off. The Mississippi baby was given drugs 30 hours after birth. Two years later, and 10 months after stopping drug treatment, there were no detectable signs of the virus and no HIV antibodies in her blood. The child is now more than three years old and remains virus-free, Deborah Persaud of the Johns Hopkins Children's Center, Maryland, who has run tests on both infants, told the meeting yesterday.

Also last year, it was announced that the "Visconti group" of 14 adults who received treatment between 35 days and 10 weeks after infection – much earlier than usual – were also virus-free after an average of seven years without medication.

Research in rhesus macaques shows why early treatment is critical to keeping the virus under control. Although the experiment involved SIV, the monkey equivalent of HIV, the results are likely to apply to the human form, say the researchers.

Afam Okoye of the Oregon National Primate Research Center in Beaverton, Oregon, and his colleagues injected 22 macaques with SIV and divided them into three groups according to their speed of treatment. The first two monkeys received antiretroviral therapy seven days after infection, the second two 10 days after, and the final 18 at 42 days later.

In all groups, viral levels in the blood peaked 12 days after treatment began. In the two monkeys treated earliest, this peaked at 71,000 copies of SIV RNA per millilitre of blood. Six weeks later, all detectable signs of the virus had disappeared.

Narrow window

In the pair of monkeys treated just three days later, viral levels peaked at 1.8 million copies of RNA per millilitre. This time it took three times as long, 18 weeks, to drive the virus down to undetectable levels.

In the remaining 18 monkeys not treated until three months after infection, viral levels peaked at 14 million copies of RNA per millilitre. But this time, even after 20 weeks of treatment, it proved impossible to eradicate all traces of the virus.

Seventeen weeks after infection, the levels of virus lurking in bone marrow, lymph nodes and gut tissue were 10 times higher, on average, in the monkeys treated at 10 rather than 7 days, even though the pairs were treated only three days apart.

These locations are important because this is where HIV is thought to establish long-lived reservoirs of dormant – non-replicating – virus. Since antiretroviral therapy works by disrupting viral replication, it has no effect on these reservoirs. This means the virus can re-emerge and re-establish infection if treatment is discontinued. This is probably the reason the virusbounced back in two men – the "Boston patients" – who initially appeared to have been purged of HIV following bone marrow transplants. The virus returned after the two had stopped drug treatment for 12 and 32 weeks respectively.

.Other AIDS researchers focusing on the quest for an outright cure are impressed by the work. "Okoye's study tells us that there's a very small window following infection in which early treatment can dramatically reduce the size of the reservoir," says Sharon Lewin of Monash University in Melbourne, Australia.