Research published this week in the journal Science failed to verify even one case of transient infection among 42 cases where infants showed evidence of HIV-1 infection contracted from their mothers, but somehow became free of the virus that causes AIDS. The researchers concluded that laboratory errors created false readings when the specimens were initially tested.
"The study of transient HIV-1 infection draws attention because of clues it might give to development of a vaccine," said principal investigator Dr. Lisa M. Frenkel, associate professor of pediatrics and laboratory medicine at the University of Washington and Children's Hospital and Regional Medical Center in Seattle. "If individuals can be identified who have transient HIV-1 infection, a study of their immune responses could give clues as to what type of immune response could protect against or eliminate the infection."
To study cases of possible transient HIV-1 infection, Frenkel and colleagues looked to exposed infants, because, in contrast to adults, the time of exposure and the source of the virus are known: from an HIV-positive mother during the birth process. This knowledge allowed blood samples taken close to the time of exposure to be studied. When virus was found, results could be verified with genetic tests.
Frenkel explains the genesis for the current research: A few years ago, she and colleagues were studying the resistance of HIV to the AIDS drug, AZT. In their research, they encountered the case of a mother who had had two positive cultures for HIV, and her baby, who had tested positive three times. During later testing, HIV could not be found in their blood specimens. "It looked as if we had a case where both the mother and infant were transiently infected," said Frenkel.
To evaluate this case, the researchers re-analyzed the test results of these two and other reported cases. In addition to the mother and child, they found another 41 cases of suspected transient HIV infection among 1,562 infants exposed to HIV in the birth process. (The others did not become infected.) The cases came from five large multi-center trials across the country. In all 43 cases, blood tests had originally shown evidence of HIV infection on one or more occasions, but later tests did not find evidence of HIV-1.
The researchers re-analyzed the infant specimens that had tested positive for HIV-1 and the mothers' specimens. They performed tests to study the genetic code in the babies' and mothers' specimens, to determine whether the virus in the baby's specimen was related to the mother's virus. "The test results in the case where both the mother and baby were positive and later showed no evidence of infection showed that none of their five specimens with HIV-1 had viruses that were related," said Frenkel. "These viruses looked genetically different enough that they did not come from the same source. Also, one of the viruses was a known strain used in laboratories."
In 20 of the other 41 cases, the researchers could not find the envelope gene -- the outside coating of the virus used to evaluate the relatedness of the genes -- which suggested that the patients' specimens had become contaminated with HIV-1 in the laboratory.
In another six cases, they found there was a mix-up in the specimens, since the infant's human gene markers did not match the gene markers found in two or more of the same infant's other virus-negative specimens.
Where they did find envelope virus in the infant specimen, they analyzed the gene sequence of the envelope virus and compared it with that of the mother. In 17 cases, they showed that the envelope virus in the child's specimen was not related to the envelope virus in the mother.
"Our hypothesis was that transient infection could occur," Frenkel said. "By using a rigorous genetic standard in examining all the supposed cases we could find from among a large number of cases, we couldn't prove transient infection in any of them. Either the specimens were mixed up, or there was contamination from one specimen to another in the laboratory."
The researchers suggest that to confirm transient infection, viral strains in the donor must be genetically linked to viral strains in the recipient. They recommend using separate laboratories to evaluate donor and recipient specimens, with reagents prepared separately by different personnel.
"Our negative studies . . . indicate that the phenomenon of transient HIV-1 infection remains to be proven and that most cases suggestive of transient HIV-1 infection are cases of mislabeling of specimens or their contamination in the laboratory," state the authors.
Co-authors include researchers from the University of Washington, U.S. Centers for Disease Control and Prevention, Medical and Health Research Association of New York City, University of Rochester, University of Texas Medical Branch, University of Minnesota, Northwestern University, University of California at Los Angeles, Aaron Diamond AIDS Research Center, Santa Fe Institute, Roche Molecular Diagnostics and New York University North Shore Hospital.
The research was supported by grants from the Pediatric AIDS Foundation, the U.S. Public Health Service and the Foster Foundation.
Journal
Science