Effect of Successive Single-gestation Pregnancies on the Course of Maternal Human Immunodeficiency Virus Disease and Perinatal Transmission

Objective: This study was undertaken to examine the effect of successive pregnancies over a 3-year period on the course of maternal human immunodeficiency virus (HIV) infection and the rate of perinatal transmission of HIV. Methods: A retrospective analysis of 32 pregnancies in 14 known HIV-infected women vs. a matched control group of HIV-infected women who had been pregnant only once was done. Results: The multiple-pregnancy group was similar to the single-pregnancy group for age, race, duration of known HIV infection, initial CD4 count, and date of first pregnancy. The delivery data were similar as well. The CD4 counts in the multiple-pregnancy group fell from 595 to 460, while counts in the single-pregnancy group fell comparably from 669 to 638, both over 37 months (P = 0.1476). Five of 5 second-born infants of known serostatus vs. 8 of 21 first-born infants were HIV-infected (P < 0.05). Conclusions: Successive pregnancies do not alter the course of HIV infection in asymptomatic women followed up to 3 years. The infants of second pregnancies of known HIV-infected women may be at higher risk for perinatal transmission.


KEY WORDS
CD4, AIDS, sterilization omen accounted for almost 15% of all new diagnoses of acquired immunodeficiency syndrome (AIDS) in the United States from June 1992 to July 1993. AIDS now represents the 5th leading cause of death in women of reproductive age. In light of this problem, multiple studies of the effect of pregnancy on human immunodeficiency virus (HIV)-infected women have been reported in recent years. While most reports in this country have concluded that pregnancy has minimal impact on the progression of HIV infection, 2-6 others have suggested an association with disease acceleration as measured by CD 4 counts or increased rates of ma-ternal infectious diseases.  In addition, minimal information exists concerning the effects of successive pregnancies on HIV-infected women, as most such patients have been advised to consider sterilization in light of the presumed 20

RESULTS
As previously stated, the individuals were case matched for race (all were African American), age, and months of follow-up after first delivery. There were 32 total pregnancies in the study group (28 live births, 4 abortions). Five women in the study group used zidovudine (ZDV) during pregnancy. Three of these discontinued the drug in the first trimester due to concern about potential adverse fetal effects. One other used dideozyinosine (DDI) and one used a blinded AIDS Clinical Trials Group (ACTG) #076 study drug. In the control group, 7 women used ZDV during pregnancy, 3 of whom discontinued the drug in the first or early second trimester. One patient in the control group used a blinded ACTG #076 study drug. Three women from the study group and 4 from the control group started antiretrovirals after pregnancy was diagnosed for maternal reasons. The timing and dosage of the drugs varied widely. The risk factors for HIV infection included patient in the study group who had a history of intravenous (IV)-drug abuse vs. 2 in the control group. There were no patients with a history of blood or blood-product transfusion; thus, the presumed mode of transmission in the remaining patients was heterosexual intercourse.
The delivery data were compared between the study and control groups, including the rates of chorioamnionitis, endometritis, low-birth-weight infants, and type of delivery (cesarean vs. vaginal) and found to be similar as well. There was case of chorioamnionitis and endometritis in both the study and control groups. Three low-birth-weight infants (<2,500 g) were born in the study group vs. 2 among the controls, and there were 5 cesarean deliveries in the study group vs. 3 among the controls.
CD 4 counts were also collected in both groups as listed in Table 1. The differences in CD 4 counts between the study and control groups were not significant at either the initial predelivery or the final postdelivery determinations (P 0.15). In addition, although mean CD 4 counts fell in the study group from 595/mm 3 to 460/mm 3 over the 37-month period examined, the control group mean fell from 669/mm to 63 8/mm at 37 months. This difference was not significant as determined by ANOVA with repeated measures (P 0. 1476). The COg percents were similar as well. Due to the wide variation in COg counts in this relatively small analysis and the fact that COg counts do not follow a standard distribution, the power in this analysis was low (power 0.06).

DISCUSSION
Multiple alterations in immune status have been reported to occur in pregnancy. [11][12][13][14] Several viral illnesses other than HIV have been reported to have increased associated morbidity in pregnancy as well. [15][16][17] The hypothesis that HIV is similarly more aggressive in pregnancy is suggested by studies showing a fall in CD 4 counts, an apparent high incidence of progression to clinical illness, and a number of pregnancy-associated deaths. 7'8'18 '9 These studies, however, are limited by either their lack of control groups, small numbers, or large numbers of IV-drug users with multiple concurrent problems being included in the study population. Relatively little information using a control group of nonpregnant infected women is available. One such recent study found minimal differences in pregnant and nonpregnant women, exam-2 ining both clinical and immunologic parameters.
A possible reason for the lack of nonpregnant control groups is that asymptomatic women at risk for infection have not consistently undergone screening for HIV until recently, in contrast to pregnant patients, who frequently have more access to testing and encouragement by health professionals. The current study deals with this problem by utilizing a control group of infected women who had been pregnant once in comparison with women who had had successive pregnancies since learning they were infected. Due to a local cultural bias against sterilization, many HIV-infected women in this community decide to maintain their fertility. More data on the long-term effects of successive pregnancies on maternal HIV status and on the risk of perinatal transmission would allow patients to make better informed decisions regarding contraception and sterilization.
These data suggest that successive pregnancies following a diagnosis of HIV infection have minimal effect on disease progression over a 3-year follow-up period in previously asymptomatic women. No opportunistic infections or AIDS-related deaths were seen; and, although differences in CD 4 counts were not significant, the low power in this analysis does not rule out the possibility of a type-II error. Furthermore, since all patients had COg counts >200, no comment can be made regarding more severely affected patients. The use of antiretroviral drugs cannot be assessed from these data. The patients using antiretroviral drugs were under the care of different physicians; therefore, the doses and timing varied. Although the use of ZDV has been shown to lessen the risk of perinatal transmission in the recently closed ACTG #076 protocol, the maternal benefits of this drug in asymptomatic patients with CD 4 counts >200 are uncertain, z The data concerning the serostatuses of the infants suggest that second-born infants are more likely to be infected than first-born infants. Although the factors affecting perinatal transmission are poorly understood, it appears that low CD 4 counts may correlate with a higher risk of transmission. Whether the relatively small drop in CD 4 counts (595 to 460) seen in the successive pregnancy group is sufficient to account for the higher transmission rate in the second-born children is unknown. Other factors suggested as possible transmission risks such as low birth weight and chorio-INFECTIOUS DISEASES IN OBSTETRICS AND GYNECOLOGY 253 amnionitis were infrequent and evenly distributed among first and second births. The cesarean delivery rate was also similar (4/21 first born vs. 1/5 second born of known serostatus).
While deserving of further analysis, the increased risk seen here should be viewed with caution due to the large number of indeterminate children and relatively small sample. The decision to delay or cease childbearing would likely be profoundly affected if the children of successive pregnancies following HIV infection were shown to be at significantly higher risk for perinatal transmission.
The current study did not attempt to document the variables in prenatal care, precise timing of maternal HIV infection, or risk behaviors in the study and control groups, which may also influence transmission. Moreover, many centers including the current study site are now routinely using ZDV in pregnant women based on the information from ACTG #076. This routine will likely decrease transmission overall, making comparison of preand post-ZDV-u,se pregnancies difficult. A prospective analysis of a sufficient number of infected women at risk for successive pregnancies would be necessary to determine the precise risk of HIV transmission to subsequent children.