1. Home>
2. Specialties>
3. HIV/AIDS>
4. Feature

Management of HIV Infection During Pregnancy: New Options, New Questions

More than a year has passed since ACTG 076 -- a clinical trial assessing the efficacy of AZT in preventing vertical transmission of HIV -- was stopped early because of a dramatic reduction in HIV infection in newborns who had received AZT from before birth. Guidelines have since been developed for the use of AZT in pregnancy, and the CDC recently recommended that all pregnant women in the United States be routinely offered HIV antibody testing.

ACTG 076 showed that transmission of HIV to the fetus can be largely prevented by treating the pregnant woman and the newborn with AZT (see ACTG 076 in Summary, p.46). This finding raises new issues for providers of primary care to women of reproductive age. This article addresses some of the questions most commonly asked about the management of pregnant women who are HIV positive or at risk for HIV.

Q: Should all pregnant women be tested for HIV infection?

A: Clearly, all women who give a history of high-risk behavior (such as IV drug use, receipt of blood products, or sex with a man who is known to be at risk for HIV infection) should be encouraged to be tested and should have the reasons for testing carefully explained.

The more difficult issue is whether to counsel testing for women who do not report such risk. While HIV infection is now one of the most common causes of death in women of childbearing age in the U.S., prevalence of the disease varies tremendously in different regions of the country and among different populations of women.

In general, HIV infection in U.S. women is most common in young women of color who reside in metropolitan areas of the eastern seaboard, in Puerto Rico, and in the rural southeast. In these areas, the chance that a pregnant woman is infected with HIV is sufficiently high that all women presenting with pregnancy should be routinely offered testing even if they do not acknowledge individual risk factors, and even if they report having had only one sexual partner. At the other extreme, in areas with very low prevalence of HIV infection in women, routine screening of women without any elicited risk factors will yield few positive tests and these will mostly be false positives. Thus, routine testing in these areas is less clearly appropriate.

However, many women without risk factors, even in low prevalence areas, ask for testing. In these cases, it is imperative that a positive test result be repeated and, where uncertainty lingers, that other tests be done, such as the viral marker assays that are currently becoming available. (See ACC May '95 for discussion of viral markers.)

Q: What should the HIV-infected pregnant woman be told about the effect of pregnancy on her own health?

A: The majority of pregnant women infected with HIV in the U.S. have CD4 counts that are well above 200 cell/mm³. For such women, current data do not suggest that pregnancy accelerates the progression of HIV infection or is associated with an increased risk of antenatal or perinatal complications in the mother. For women in more advanced stages of HIV disease, however, many experts are concerned that the added stress of pregnancy may lead to worse short- and long-term outcomes for the women and possibly for their babies. Moreover, such women are at risk for opportunistic infections (OIs), which in virtually all cases are life-threatening. They need to understand that if they do develop such complications, they will need treatment with medications that, in general, have not been studied in pregnancy, and that could be teratogenic.

There are some non-medical issues that are very important to discuss with HIV-positive patients who are making decisions about becoming pregnant or continuing a pregnancy. One is the individual patient's feelings and beliefs about bearing a child who may be HIV-infected even if the mother receives AZT during pregnancy. (In ACTG 076 there was still an 8% vertical transmission rate for treated babies.) Another possibility the mother should consider is that she may not survive long enough to raise the child to adulthood. However, it is also important to convey to the patient that there is a great deal about this infection that remains unknown, and that the prognosis for both infected children and their mothers could improve substantially in the coming years if newer, better antiviral agents and treatment strategies are identified.

Q:How late in pregnancy can AZT be given and still prevent vertical transmission?

A: While the definitive answer to this question is unknown, many experts recommend that pregnant women be offered AZT whenever they present during pregnancy after 14 weeks gestation, even if this means giving the initial AZT dose (intravenously) during labor. For children born of HIV-infected mothers who did not receive AZT, current CDC guidelines recommend treatment of the baby for 6 weeks, the same treatment as is recommended for babies whose mothers did receive the drug during pregnancy, if therapy can be started within 24 hours of birth. While AZT therapy restricted to the neonatal period is less likely to be beneficial, 076 showed no serious morbidity associated with use of AZT in this setting and thus the possible benefit is still felt to outweigh any risk.

Q: Should women receiving AZT when they become pregnant stop the drug during the first trimester?

A: Experts are divided on the question of whether to interrupt AZT temporarily in this situation. While no teratogenic effects with AZT therapy have been found, the number of women who have received the drug during the first trimester is small, and less common teratogenic effects may thus not yet have been detected. Moreover, the dramatic reduction in transmission of HIV seen in ACTG 076 occurred even though women could not enter the trial until after the 14th week of gestation, and the median time of entry was 26 weeks. However, there is a theoretical concern that viral replication could "rebound" after discontinuation of AZT and that this might have a negative effect on both mother and fetus. Although there are currently no data to support this theory, if AZT therapy is interrupted, it would seem prudent to restart it promptly after the 12th week of gestation.

Q: Should all HIV-infected pregnant women be offered treatment with AZT to prevent perinatal transmission, or only those in certain CD4 ranges?

A: ACTG 076 enrolled only women with CD4 counts above 200 because it was felt to be unethical to randomize women with low CD4 counts to a placebo-controlled trial, given the proven efficacy of AZT for prolonging survival in nonpregnant adults with CD4 counts in this range. However, it is believed likely that AZT would have at least as great an impact on vertical transmission in women with counts below 200 as in the population enrolled in ACTG 076, whose median CD4 count was closer to 500. Also, vertical transmission can probably occur in women with high CD4 cell counts, albeit perhaps at a lower rate. Most experts, therefore, recommend offering AZT to all pregnant women with HIV infection, regardless of CD4 count, to prevent vertical transmission.

Q: If a woman had a CD4 count greater than 500 before using AZT during pregnancy, should she continue to take AZT after she delivers her baby or should it then be stopped?

A: In general, women who have a CD4 count above 500 throughout pregnancy should stop taking AZT after delivery, because there is currently no evidence that AZT is beneficial to nonpregnant adults with CD4 counts in this range.

Q: How should women be managed who have had prolonged use of AZT before pregnancy, or whose viral isolate has been shown in vitro to have a high level of resistance to AZT? Should they be treated with another antiviral -- such as ddI, ddC, or d4T -- to try to prevent perinatal transmission?

A: Unfortunately, there are currently no good data to guide the clinician in this situation, although studies designed to address this are starting. One option is to enroll such patients in these new trials if they are nearby (call 1-800-TRIALS A for sites). This will provide close monitoring of the woman and her fetus and could help expand our knowledge base for future patients. If enrollment in a trial is not possible or desired, and the woman's virus is not known to be resistant to AZT, then AZT could be continued in an effort to reduce perinatal transmission. Some experts would begin an alternative agent after the first trimester, using doses shown to be safe and effective in nonpregnant adults. Another strategy would be to treat with both AZT and an additional nucleoside analogue. If possible, consultation with an experienced HIV specialist should be obtained in this situation.

Q: Should prophylactic therapy against such opportunistic infections as PCP, toxoplasmosis, MAC, CMV, and fungal disease be given to pregnant women?

A: PCP and Toxoplasmosis -- There is clear consensus among the experts that asymptomatic HIV-positive pregnant women should receive prophylaxis against PCP if their CD4 count is below 200, and prophylaxis is in order up to a count of 250 if the patient also has thrush, recurrent vaginal candidiasis, or another manifestation of HIV infection. TMP-SMX (Bactrim, Septra), dapsone, and aerosolized pentamidine are all thought to be safe in pregnancy, although some experts would avoid dapsone in the first trimester and some do not like to use TMP-SMX close to term because of the theoretical risk of kernicterus in the newborn. However, of the three agents, TMP-SMX has the greatest efficacy against PCP and most experts recommend its use in all pregnant women who need PCP prophylaxis. It is also thought to provide prophylaxis against toxoplasmosis. Thus, if the patient can tolerate the drug, TMP-SMX should be used in pregnant HIV-infected women, especially if they have a positive toxoplasma serology.

MAC -- Rifabutin is now commonly used to provide primary prophylaxis against Mycobacterium avium intracellulare (MAC). It has been shown to reduce the incidence of MAC bacteremia, fever, and abnormal liver function tests, although it has not yet been definitively demonstrated to reduce the incidence of invasive MAC disease or to improve survival. It is recommended for use in asymptomatic HIV-infected persons with CD4 counts below 50 or in people with fewer than 100 CD4 cells who also have had an AIDS-defining opportunistic infection such as PCP.

There are essentially no data on the use of rifabutin during pregnancy, so its teratogenic potential is unknown. While consensus on rifabutin in pregnancy is lacking, the risk of MAC is very high in patients with advanced AIDS. Once it occurs, MAC necessitates the use of multiple toxic agents and, untreated, usually leads to death within a year of diagnosis. Moreover, the negative consequences of MAC on pregnancy are likely to be severe, given its constellation of high spiking fevers, weight loss, and other constitutional manifestations. For these reasons, pregnant HIV-infected patients should be strongly considered as candidates for rifabutin therapy after the twelfth week of pregnancy if their CD4 counts are less than 50, or less than 100 if there is a prior AIDS-defining diagnosis.

CMV -- A recent clinical trial in nonpregnant adults has demonstrated that, compared with placebo, primary prophylaxis with oral ganciclovir significantly delayed the onset of CMV retinitis and colitis in patients with advanced HIV infection, although it did not extend survival. There is not yet consensus on the cost/benefit ratio of using this agent for primary prophylaxis. Oral ganciclovir caused bone marrow suppression severe enough to require the use of G-CSF and erythropoietin in a substantial proportion of patients. The cost of a year of treatment with oral ganciclovir has been estimated to approach $15,000 (not including the use of colony-stimulating factors). Moreover, there are no data on teratogenicity. Given these concerns, there does not appear to be sufficient reason to recommend the routine use of this agent in CMV-positive pregnant women. An alternative strategy is to do frequent eye examinations in this population. Although there is no experience with ganciclovir use in pregnancy, women who are documented to have life-threatening or sight-threatening CMV disease during pregnancy should be offered aggressive treatment with IV ganciclovir or foscarnet.

Fungal disease -- While oral fluconazole has been demonstrated to significantly reduce the incidence of systemic fungal disease in patients with AIDS, most experts argue against its use as a primary prophylactic agent because of concerns about the emergence of fluconazole-resistant candida. In addition, fluconazole may exert an anti-estrogen effect of uncertain significance. Thus, there is no reason to recommend its use in pregnancy for primary prophylaxis. In addition, most experts try to avoid fluconazole as treatment for vaginal candidiasis unless it is unresponsive to or recurrent after repeated treatment with an intravaginal topical antifungal agent. In pregnant patients requiring maintenance therapy for cryptococcal meningitis or other systemic mycoses, amphotericin should be used in the first trimester, since there is more experience with this drug in pregnancy than with fluconazole. After the twelfth week of pregnancy, fluconazole is indicated, as it is superior to amphotericin for maintenance therapy in this setting.

Q: In deciding when to initiate prophylaxis for opportunistic infections in pregnant women, should the usual CD4 cutoffs be used?

A: CD4 counts decline somewhat during pregnancy in non-HIV-infected women, usually reaching a nadir in the third trimester and returning to normal levels by six weeks postpartum.

In HIV-infected women, this drop may be somewhat more pronounced and CD4 counts may not return to normal after delivery -- unsurprisingly, given the steady loss of CD4 cells over time in most HIV-infected people. At present, however, there are no data to suggest that pregnant HIV-infected women become at risk for OIs at a different level of CD4 cells than nonpregnant HIV-infected adults. Thus the usual cutoffs should be used and the clinician should respond to a given CD4 count in a pregnant woman in the same way as in a nonpregnant patient.

Q: What about women whose serologies for toxoplasma and CMV are negative? What is the risk of primary infection with these agents in HIV-infected pregnant women? How can it be lessened?

A: Primary infection with both toxoplasmosis and CMV can be devastating for both the pregnant HIV-infected woman and her fetus. Patients should be strongly counselled to avoid eating raw or undercooked meats and to avoid contact with the feces of cats, especially kittens. They should not empty litter boxes or, if there is nobody who can do this for them, should wear gloves, wash their hands well afterward, and be exceedingly cautious. Good handwashing is the best prevention for CMV and should be emphasized with pregnant HIV-infected women, especially those who are healthcare workers or who have children in day care.

Q: Should labor and delivery be managed differently for women with HIV infection?

A: Few data are available to guide clinicians on managing labor and delivery in HIV-infected women. Scalp electrode monitoring should be avoided when possible. Premature rupture of membranes increases the risk of fetal infection. The possible benefit of Cesarean section is unknown. It has been shown that in twin pregnancies of HIV-infected mothers, the first twin is more likely to be infected, suggesting that prolonged exposure to maternal blood and secretions may increase risk. However, Cesarean section has associated maternal morbidity and even mortality. Thus, at present, most obstetricians advise performing sections in HIV-infected women only for the same indications as in non-infected patients.

Q: Since 75% of babies born to HIV-infected women aren't going to become infected even without AZT, is the benefit of AZT really greater than the risk of unknown long-term side effects?

A: While it is true that the long-term effects of AZT on both infected and uninfected children are unknown, a two-thirds reduction in the incidence of a fatal disease is very dramatic and makes it unlikely that later, long-term side effects will outweigh this benefit. It is not yet possible to determine antenatally which babies are most likely to be infected; if this can eventually be done, then AZT therapy could target these babies, improving the risk-benefit ratio even more.

Providers who treat women with AZT or other antiretrovirals during pregnancy should contact the following registry to complete a brief, anonymous report on pregnancy outcome. Registrar, Antiretroviral Pregnancy Registry, P.O. Box 12700, Research Triangle Park, N.C., 27709-2700. Telephone 800-722-9292, Ext. 58465.

— Deborah Cotton, MD, MPH, and Heather Watts, MD

Dr. Cotton is Editor of AIDS Clinical Care. Dr. Watts is Associate Professor of Perinatal Medicine, Department of Obstetrics and Gynecology, University of Washington Medical Center, Seattle.

Published in AIDS Clinical Care June 1, 1995

Citation(s):

Connor EM et al. Reduction of maternal-infant transmission of human immunodeficiency virus type 1 with zidovudine treatment. N Engl J Med 1994 Nov 3 331 1173-1180.

• Original article (Subscription may be required)
• Medline abstract (Free)

Rogers MF and Jaffe HW. Reducing the risk of maternal-infant transmission of HIV: A door is opened. N Engl J Med 1994 Nov 3 331 1222-1223.

• Original article (Subscription may be required)
• Medline abstract (Free)

Bayer R. Ethical challenges posed by zidovudine treatment to reduce vertical transmission of HIV. N Engl J Med 1994 Nov 3 331 1223-1225.

• Original article (Subscription may be required)
• Medline abstract (Free)

Centers for Disease Control and Prevention. Recommendations of the U.S. Public Health Service task force on the use of zidovudine to reduce perinatal transmission of human immunodeficiency virus. MMWR Morb Mortal Wkly Rep 1994 Aug 5 43 RR-11 1-20.

Cotton D. AIDS in women. In Broder S et al. Textbook of AIDS Medicine. Williams and Wilkins, Baltimore 1994 161-168.

AIDS Clinical Care 1994 Dec 104-.

AIDS Clinical Care 1994 Aug 69-.

AIDS Clinical Care 1994 Apr 33-.