Baby's AIDS virus infection vanishes.

Why do some 75 percent of infants born to women with the AIDS virus fail to acquire their mother's infection during her pregnancy? A provocative new report suggests one possible answer: Many of the children may have hosted the virus briefly, only to eventually clear it from their systems.

There have been sporadic reports of infants who showed signs of HIV, the AIDS-causing virus, at birth but failed to exhibit the virus when tested again, months later. However, many AIDS researchers dismissed such accounts, arguing that the initial positive report likely resulted from laboratory error.

Now researchers at the University of California, Los Angeles, School of Medicine report unambiguous evidence of a boy who tested positive for HIV twice -- at 19 days of age and 1 month later. Yet by every measure, this kindergartner appears to have been HIV-free for at least 4 years, report Yvonne J. Bryson and her colleagues in the March 30 New England Journal of Medicine.

To rule out the possibility of lab errors, her team analyzed DNA sequences from the protein envelope around the HIV that researchers had cultured from the infant's blood during both the early tests. For all practical purposes, the two samples of virus appeared identical, the UCLA team notes.

Because the hospital never kept samples of the mother's blood from the time of her son's delivery, the researchers attempted to match the infant's HIV to the virus circulating in his mother 1 year later. And even though the virus can mutate quickly inside its host, the boy's HIV matched one of the forms that his mother carried.

Overall, Bryson's team concludes, when the DNA sequencing and genetic typing of the HIV samples are taken together, "it is highly likely that [the virus] found in the infant originated from his mother." Moreover, they add, the fact that the child carried the virus for almost 2 months suggests his infection had been active.

In an accompanying editorial, Kenneth McIntosh and Sandra K. Burchett of Children's Hospital in Boston argue that the current report suggests that at least some previous reports of HIV clearance in infants were correct. If so, they add, it may prove possible to design an immunology-based approach to thwart HIV's transmission from mother to baby.