The use of zidovudine in pregnant patients infected with human immunodeficiency virus type 1 (HIV-1) has dramatically decreased the perinatal transmission rate of HIV-1 from 30 percent to less than 10 percent. The mechanisms by which zidovudine prevents transmission are not fully understood, but at least one proven benefit of antiviral therapy is a decrease in the maternal level of HIV-1 RNA. Garcia and colleagues, who are part of the Women and Infants Transmission Study (WITS), prospectively evaluated the hypothesis that the higher the maternal HIV-1 RNA level, the more likely the mother is to transmit the virus to her infant. They also attempted to determine if high HIV-1 RNA levels early in pregnancy correlated with prenatal transmission of the virus.

Women and their infants have been enrolled in WITS since 1989. The patients have been followed at eight university hospital centers. The women included in the HIV-RNA component of this study were enrolled between 1990 and 1995 and gave birth to singleton infants. Patient evaluations were performed at least three times during the pregnancy and at delivery (not including routine prenatal care). At each visit, an HIV-1 RNA level and CD4 lymphocyte count were determined. During the course of this study (after March 1994), maternal zidovudine therapy became the standard of care. Women who received zidovudine were classified separately for purposes of data analysis. Infants born to the women in the study had blood obtained for viral culture within the first six days of life, followed by repeat testing at 1, 2, 4, 6, 9, 12, and 18 months of age, and every six months thereafter. Infants who had two or more blood cultures positive for HIV-1 were considered positive. After April 1994, blood cultures from infants were additionally obtained within the first 48 hours. Infants whose cultures were positive within the first 48 hours were considered to have acquired the infection in utero. Infants whose cultures were positive at a later time were considered to have been infected during labor and delivery.

The study included 552 women; average age was 28 years, and more than 80 percent were black or Hispanic. The overall rate of HIV-1 transmission to the infants was 20.6 percent. Before March 1994 (prezidovudine), the rate was 24 percent; afterward, it was only 9 percent. In evaluating the rate of perinatal transmission as it correlated with HIV-1 RNA levels, none of 57 women whose RNA levels were less than 1,000 copies per mL transmitted HIV-1 to their infants—whether or not they received zidovudine. In women whose RNA levels were between 1,000 and 10,000 copies per mL, the rate of transmission was 16.6 percent. The transmission rate increased to 21.3 percent in women with RNA levels of 10,000 to 50,000 copies per mL and to 30.9 percent in women with RNA levels between 50,000 and 100,000 copies per mL. The rate was 40.6 percent if the RNA level was greater than 100,000 copies per mL; if these women did not take zidovudine, 63.3 percent of their infants became HIV-infected.

Sixteen of the infected infants had a blood culture result positive for HIV-1 within 48 hours of birth. However, no difference in the median levels of maternal HIV-1 RNA was found in their mothers compared with the mothers whose children were found to have later (presumed intrapartum) infection. This finding would indicate that there is no relation between maternal HIV-1 RNA levels at any time during pregnancy and the risk of either early or late perinatal transmission.

The findings of this study suggest that the maternal level of HIV-1 RNA best predicts the risk of perinatal transmission of the virus. Perinatal transmission occurred independently of CD4 count, although an association with CD4 percentage and absolute CD4 count to the rate of viral transmission was shown.

editor's note: A second study published in the same issue of the journal showed similar findings. These data should further encourage the use of antiretroviral therapy in all HIV-infected women during pregnancy to maintain as low an HIV-1 RNA level as possible. More than 80 percent of the patients in the first study delivered vaginally, which suggests that if the maternal HIV-1 RNA level is less than 1,000 copies per mL, especially close to the woman's due date, the use of elective cesarean section (as has been suggested by a recent meta-analysis of other studies), is unnecessary.—j.t.k.