Background:  It is estimated that 25 to 30% of HIV-infected infants will progress rapidly to AIDS and death in the first year of life. Serologic tests for HIV infection do not differentiate between exposure and infection due to maternally acquired antibodies. Virologic tests like DNA or RNA polymerase chain reaction (PCR) are confirmatory but difficult to perform in resource-constrained settings. Our aim was to evaluate whether the CD4 count, CD4 percentage, or CD4/CD8 ratio could serve as a surrogate marker for HIV infection in infants under 18 months of age.

Methods:  We studied 88 infants whose mean age was 4.0 months (range 0.2 to 18); 47 were female, 41 males. They were all born to HIV⁺ mothers and were referred to the Tuberculosis Research Centre, Chennai, between January and August 2006. DNA PCR was performed using the Roche Amplicor HIV-1 DNA Test Version 1.5 qualitative kit and CD4 and CD8 counts determined by 2-color flow cytometry (BD FACS caliber).

Results:  While 66 infants were negative, 22 were DNA PCR positive. There were significant differences in mean CD4 percentage, CD4 count, and CD4/CD8 ratio between HIV-infected and -uninfected infants. Distribution of CD4 counts in HIV⁺ and HIV^– infants is shown in the table. The mean CD4 percentage and counts of PCR-positive infants were significantly lower than those of the PCR-positive infants. The CD4/CD8 ratio was <1.0 in all 22 infected children and in 4 of the 66 uninfected children (sensitivity 100%, specificity 94%).

Conclusions:  Our findings suggest that the CD4/CD8 ratio may be used as a sensitive surrogate marker of HIV infection in an infant born to a HIV⁺ woman, since CD4/CD8 counting facilities are more widely available than virologic assays, in resource-poor settings. This would help identify infants for cotrimoxazole prophylaxis and ART.

* p <0.001