Reliability of laboratory markers of HIV-1 infection in Argentinian infants at risk of perinatal infection.

Early and accurate diagnosis of HIV-1 infection in infants born to HIV-1-seropositive mothers is of great importance. Polymerase chain reaction (PCR), HIV culture, and p24 antigen detection assays were evaluated for their ability to detect the presence of HIV in 195 infants at risk of perinatal infection. Using the Centers for Disease Control and Prevention guidelines for assessing HIV infection status in children younger than 18 months, 70 infants (36%) were diagnosed as HIV-1 infected and 125 (64%) lacked virologic and clinical evidence of infection. PCR and HIV culture were the most sensitive laboratory markers, detecting 100% and 98% of positive samples, respectively, regardless of age at testing. HIV-1 p24 antigen assay was detected in 26 of 38 positive samples but not in negative samples. PCR was performed with three different sets of primers (SK38/SK39-SK19-gag, SK68/SK69-SK70-env, and SK150/SK431-SK102-gag). The sensitivity/specificity of the individual assays were for SK19, 96.1%/94.25%; SK70, 89.6%/100%; and SK102, 100%/100%. A sample was considered HIV-1 positive when two positive PCR results were obtained with two different pairs of primers, and negative if the sample was negative when three sets of primers were used. False-positive results were occasionally obtained with probe SK19 in six seroreverter infants before serologic status was known. This suggested that the infection was caused by nonreplicative strains or were false-positive results probably by nonspecific amplification due to cross-reaction with other microorganisms; contamination was discarded because there was no specific amplification with the other two primers. All the HIV-1-infected infants were correctly identified with PCR; all except one could be identified with coculture and only 68.4% were confirmed with p24 antigen assay. No seroreverter infant was misdiagnosed using the criteria selected.