As demonstrated by ACTG 076, a three-part course of AZT treatment reduces the rate of vertical transmission by 67%. However, the role of the mode of delivery has been less clear. Since a large proportion of vertical transmission occurs at or near delivery, any intervention at this time might prove beneficial. Elective Cesarean sections prior to the rupture of membranes can prevent the infant from being exposed to maternal blood and secretions while passing through the birth canal. But studies attempting to demonstrate the effectiveness of such C-sections in reducing vertical transmission have been inconclusive due to various confounding factors such as concurrent use of antiretroviral therapy and prior rupture of uterine membranes. Presentations at Geneva revealed the most convincing data to date on enhanced reduction in vertical transmission in women who were both on antiretroviral treatment and chose elective C-sections before their membranes ruptured.
Three European observational studies provided the first data: Laurent Mandelbrot, M.D., reported the results of the largest and most impressive of these (abstract 23272). Among 902 women in the French Perinatal Cohort who were treated with AZT, elective C-section resulted in a rate of vertical transmission of only 0.8%. In comparison, emergency C-sections had an 11.4% rate of transmission, and the rate for normal vaginal delivery was 6.6%. Christian Kind, M.D., of the Swiss Neonatal Group presented data from an ongoing nationwide prospective study (abstract 23599).
Among 45 women who completed the full 076 regimen and had elective C-section, there were no cases of vertical transmission. Finally, Renate Lutz-Friedrich, M.D., reported on the German Perinatal Cohort study of 255 mother-child pairs (abstract 23291). Since 1994, elective C-section was performed in addition to AZT treatment in 80 HIV-positive pregnant women. The rate of transmission in this group was reduced to 2.5%. For women who had elective C-sections but no AZT, the transmission rate was 10.8%. Transmission was 7% in AZT-treated women who had vaginal deliveries.
Augusto Semprini, M.D., of Milan, Italy, presented the results of a five-year international trial (abstract 23599). Pregnant HIV-positive women with similar antiretroviral regimens were randomized to either elective C-section at 38 weeks or spontaneous vaginal delivery. Of the 133 children delivered by C-section, 3% contracted HIV compared to 10.3% of 132 infants delivered vaginally. Maternal viral load data is still under analysis.
The largest survey by far was conducted by the NIH-funded International Perinatal Group, which performed a meta-analysis of data from five European and ten North American prospective studies (abstract 23603). The primary analysis included data from 8,533 mother-child pairs. After adjusting for receipt of antiretroviral therapy, maternal disease progression and birth weight, risk of vertical transmission was reduced by over 50% with elective C-section compared to other modes of delivery. In women who received antiretroviral therapy, the rate of transmission was 2% with elective C-section and 7.3% with other modes of delivery. In women who were not on antiretroviral therapy, the transmission rates were 10.4% with elective C-section and 19% with other modes of delivery.
The enhanced protection of an elective C-section must be weighed against the risk of postoperative complications. Lynne Mofenson, M.D., of NIH, pointed out that the benefit of an elective C-section might be minimal compared to the risk of surgery for a pregnant woman who is taking anti-HIV medications and has a viral load below the level of detection and a CD4 cell count of 500. She stressed that rather than making a universal recommendation, each case needs individual assessment.
Beyond AZT, there is not much information from clinical trials to help an HIV-positive pregnant woman make treatment decisions. While using combination therapy may be best for her own health, she will also be concerned about the safety of her infant. In Geneva, there were some presentations that addressed this topic.
Karen Beckerman, M.D., of the Bay Area Perinatal AIDS Center (BAPAC) at San Francisco General Hospital, reported on changing treatment trends over the past three years in more than 200 HIV-positive women (abstract 12151). Choice of treatment has shifted from AZT monotherapy in 1996, to mostly dual therapy and some triple therapy in 1997, to mostly triple therapy in 1998. Combination therapy has been well tolerated and brought maternal viral load to below 400 copies/ml in most cases. No maternal or fetal complications have been observed. Of the 60 infants born since 1995, 43 are uninfected and 17 are less than six months old but have so far tested negative by the PCR viral load test.
Alice Stek, M.D., of the Los Angeles County-University of Southern California Medical Center, presented the results of a case review of 14 HIV-positive pregnant women who received nevirapine, AZT and another nucleoside analog (nine ddC, two 3TC and one ddI later switched to 3TC) from June 1997 to January 1998 (abstract 12152). The regimens have been well-tolerated in the women, with three rashes and one case of AZT-induced anemia. In 82% of cases, viral load was sustained below 200 copies/ml. Eight participants have delivered and seven of the newborns test negative for HIV-infected cells (by PCR assay for HIV DNA). Test results for the one remaining baby are pending. All infants were born without abnormalities.
Results were not as positive in a Swiss cohort study, presented as a late breaker by Patrizio Lorenzi, M.D., of the Geneva University Hospital (LB 10/abstract 32453). Between October 1996 and May 1998, 37 HIV-positive pregnant women were evaluated. Forty-three percent received a protease inhibitor-containing regimen. The rest were taking two nucleoside analogs. Twelve women became pregnant while on treatment (five on protease inhibitors). All were on treatment at delivery. Maternal viral load at delivery was below 400 copies/ml in 18 of 30 women.
The most common side effect in the mothers was anemia, and they experienced no unexpected or life-threatening adverse events. Among the 30 infants born to date, 33% were born prematurely (before 37 weeks). This compares to 15% to 17% of babies born to mothers in a previous study who had received AZT monotherapy or no treatment. There were three serious adverse events among the newborns: two infants experienced non-life-threatening cerebral hemorrhages and one had a rare birth defect that caused a malformation in the bile tract. The rate of vertical transmission was low with one HIV-positive infant identified. This baby was delivered vaginally and the mother did not adhere to her antiretroviral regimen.
The sample size of this cohort was small, some of the study subjects were IV drug users (which contributes to premature births), and the comparison group was from a previous study of women who might have had different characteristics. Dr. Mofenson pointed out that there was no adjustment for two other possible confounding factors: "It is critical to control for the CD4 count and disease stage of the mother. If those two things are important risk factors for problems at birth, and women on combination therapy treatment are more likely to have lower CD4 counts and more advanced disease stage, then the increase in premature births may be because the women are sicker as opposed to the antiretroviral drugs causing the problems."
Nonetheless, the results of the Swiss study underline the potential risks to the newborn of combination therapy. Four NIH-sponsored ACTG trials on protease inhibitor use in pregnant HIV-positive women recently were put on temporary hold after it was determined that out of 11 infants born thus far, four were delivered prematurely. One developed an infection and died a week after birth. The other premature infants are healthy. One full-term infant died of respiratory failure and there was one stillbirth at 22 weeks.
It is not clear if the premature births in these trials are due to antiretroviral therapy, severity of maternal HIV disease or other risk factors. ACTG investigators analyzed data on 3,500 women from various cohort studies in the U.S. and preliminary results indicated that women who received no therapy had the highest rate of premature births (33%). In women who received monotherapy or combination therapy (with or without a protease inhibitor), rates of premature delivery ranged from 19% to 25%. (The overall rate of premature delivery is 11% across the country.) These data suggest that the premature births seen in the ACTG study may be due more to the presence of HIV disease than to antiretroviral therapy. Ongoing reviews are evaluating how maternal factors may contribute to premature births.
Protocol teams are now amending the entry criteria to exclude women at high risk for premature delivery, and Dr. Mofenson expects the ACTG trials to reopen again within the next few weeks. In the meantime, NIH officials recommend that pregnant women on combination therapy for HIV infection continue their regimens as prescribed by their physicians.