Recommendations for Use of Antiretroviral Drugs in Pregnant Women Infected With HIV-1 for Maternal Health and for Reducing Perinatal HIV-1 Transmission in the U.S.

FAQ answers by Lynne Mofenson, M.D. 5/22/01

Q: The Uganda study seems to show that nevirapine is the best drug for intrapartum prophylaxis. The guidelines actually offer two other regimens for use when only intrapartum/postpartum prophylaxis can be given, but nevirapine would appear to be the best.

A: The guidelines currently recommend a choice of four intrapartum/postpartum regimens for use when HIV-infected pregnant women have not received antenatal antiretroviral therapy. These include: 1) single dose nevirapine at the onset of labor followed by a single dose of nevirapine for the newborn at age 48 hours, recommended based on the results of HIVNET 012; 2) oral AZT and 3TC during labor followed by one week of oral AZT/3TC for the newborn, recommended based on the results of the PETRA trial; 3) intrapartum intravenous AZT infusion followed by six weeks of AZT to the newborn, based on epidemiologic data; and 4) the two-dose nevirapine regimen combined with intrapartum intravenous AZT and six weeks of AZT to the newborn, based on hypothetical but unproven enhanced benefit of the combination.

It is not possible to directly compare efficacy between the trials using the recommended intrapartum/postpartum regimens because the studies were conducted in different populations and had different comparison groups. The HIVNET 012 nevirapine trial was conducted in breastfeeding women and the nevirapine regimen was compared to an intrapartum/one week infant AZT-only regimen. Efficacy data are now available through age 12 months, with a persistent reduction in transmission of 41% (95% confidence interval (CI) 16-59%) with the nevirapine compared to ultra-short AZT regimen (Owar M. XIII International AIDS Conference, Durban, South Africa, July 2000, Abstract LbOr1). The AZT/3TC PETRA trial was conducted in a mixed population of breast- and formula-feeding women and was compared to placebo. While 6-week data on infection status showed ~36% efficacy of AZT/3TC in reducing transmission, data on HIV infection or death did not demonstrate persistent efficacy through age 18 months (7% lower infection rate in the AZT/3TC than placebo group) (Gray G. XIII International AIDS Conference, Durban, South Africa, July 2000, Abstract LbO5). The intravenous intrapartum/6 week infant AZT regimen has not been evaluated in a clinical trial, but was evaluated in an epidemiologic study in non-breastfeeding women; in this study, the comparison group was women not receiving AZT, and was subject to potential selection bias (Wade N. N Engl J Med 1998;339:1409-14). There was a 38% decrease (95% CI, 18-81%) in the risk of infection with intrapartum/ postpartum AZT compared to no AZT.

The SAINT trial was conducted in South Africa directly compared the two-dose nevirapine regimen (with the addition of a second single dose to mother at time of the infant single dose at age 48 hours due to breastfeeding) to the intrapartum/one week postpartum AZT/3TC PETRA regimen; approximately 40% of women breastfed. Results of this study were presented at the XIII International AIDS meeting (Moodley D. XIII International AIDS Conference, Durban, South Africa, July 2000, Abstract LbOr2). At age 2 months, infection rates were 13.3% in the nevirapine arm and 10.2% in the AZT/3TC arm, which were not significantly different. Both regimens were safe and well-tolerated, and no serious hepatic toxicity or rashes were observed in mothers or infants. Nevirapine offers the benefit of minimal cost and ease in administration (and could include directly observed therapy) compared to the other possible regimens.

However, nevirapine resistance can be induced with a single dose of nevirapine. In a study of 95 women who received nevirapine in HIVNET 012, nevirapine resistance mutations (primarily K103N mutation) were detected in 17/95 (18%) of women (Eshleman SH. 8th Conference on Retroviruses and Opportunistic Infections, Chicago, IL, February 2001, Abstract 516). Development of resistance correlated with higher viral load, lower CD4+ cell count, and longer drug half-life, suggesting that resistance was increased with prolonged exposure to subtherapeutic levels of nevirapine in the presence of replicating virus. Samples were available from 5/17 women with resistance mutations at 12-18 months after delivery, and no detectable resistance was found at that time. The clinical significance of this finding is unknown; there was no association of nevirapine resistance and risk for maternal death. While nevirapine resistant virus fades from detection, with wild-type virus again predominating after drug selective pressure is removed, it is unknown whether this finding will compromise future use of non-nucleoside reverse transcriptase inhibitors for treatment in the women. These data suggest the importance of initiating HAART (HAART) in the immediate postpartum period in women who receive nevirapine prophylaxis in the US.

Q: Is there consideration for nevirapine for the intrapartum component of the standard regimen?

A: A perinatal trial, PACTG 316, conducted in the U.S., Europe, Bahamas and Brazil, attempted to address this question (Dorenbaum A. 8th Conference on Retroviruses and Opportunistic Infections, Chicago, IL, February 2001, Abstract LB7). The study compared the addition of the two-dose nevirapine regimen versus placebo to standard antiretroviral prophylaxis and therapy; 23% of women were receiving AZT alone, 28% AZT/3TC, 8% other combination without protease inhibitors, and 41% combination with protease inhibitors. Delivery HIV RNA was <400 copies/ml in 49% of women. The overall transmission rate in the study was only 1.5%, and did not differ between nevirapine and placebo arms (1.5 vs 1.4%, respectively). No significant differences were observed between treatment arms in any subgroup, including viral load and type of antenatal therapy. The low transmission rate precluded definitive conclusions related to the additional efficacy of the two-dose nevirapine regimen. However, one can conclude among HIV-infected women receiving prenatal care and who are treated with standard antiretroviral prophylaxis (generally combination therapy), the risk of perinatal transmission is low and the addition of the two-dose nevirapine regimen did not offer clinically significant benefit in further reducing transmission.

Additionally, development of nevirapine resistance was observed in women receiving the two-dose nevirapine regimen despite the receipt of antiretroviral drugs antenatally. Genotyping was performed on 46 women in PACTG 316 from the U.S. with available samples who received nevirapine and had delivery RNA >400 copies/ml; 5/46 (11%, 95% CI 3.6-23.6%) of women developed a new nevirapine resistance mutation (primarily the K103N mutation) at 6 weeks postpartum (Cunningham C. 8th Conference on Retroviruses and Opportunistic Infections, Chicago, IL, February 2001, Abstract 712). In a similar study among the French cohort enrolled in PACTG 316 who had HIV RNA >200 copies/ml at delivery, nevirapine resistance mutations were detected at 6 weeks postpartum in 7/43 women (10%) (Chaix ML. 8th Conference on Retroviruses and Opportunistic Infections, Chicago, IL, February 2001, Abstract 470). Viral load at delivery ranged between 308-11,585 copies/ml in the women who developed resistance. The addition of the two-dose nevirapine regimen to standard antiretroviral therapy therefore carries a potential risk of development of nevirapine resistance in women with detectable viral replication, and appears to have no clinically significant impact in lowering further transmission.

Q: The recommendations seem to include AZT even when there is virologic failure. Is it better to continue AZT or to change to an alternative NRTI combination to achieve virologic success?

A: Treatment of HIV-infected pregnant women should be based on the same principles as treatment of non-pregnant individuals. In the case of a non-pregnant adult, development of virologic failure would require a change in therapy, including use of at least two new antiretroviral agents and optimally use of all new agents; this would be the same for a pregnant woman. When possible, AZT should be a part of the regimen. However, in circumstances where AZT would not be advisable (e.g., if d4T is a component of the new therapeutic regimen or intolerance to the drug), the woman should receive a therapeutic regimen that is optimal for her own health. It is recommended that intrapartum intravenous AZT and the 6-week infant AZT prophylaxis regimen should be administered even if antenatal AZT was not given.

There are likely several mechanisms by which antiretroviral therapy prevents transmission. The most important mechanisms are likely reduction in maternal blood and genital viral load and pre-exposure prophylaxis of the infant, through passage of the drug across the placenta to the fetus prior to intensive exposure to HIV in maternal genital secretions during birth. An optimal therapeutic regimen that lowers viral load could provide benefit in preventing transmission. However, in addition to lowering viral load, the ability of at least one drug in a treatment regimen to cross the placenta and reach virucidal levels in the fetus is an important component of a prophylactic regimen. Among the NRTIs, AZT and 3TC have the greatest placental transfer, followed by d4T; ddI and ddC have substantially less placental transfer than the other NRTIs. Since there are some concerns regarding teratogenicity of delavirdine and efavirenz in animals and these drugs have not been studied in pregnant women, nevirapine is the only current NNRTI that might be considered for use in pregnancy. The available data suggest that protease inhibitors have very little placental passage. AZT may be unique among the NRTIs in that it is phosphorylated in the placenta, whereas the other NRTIs studied to date are not; thus, AZT could provide particular protection against intrauterine transmission. Therefore, whenever possible, it is recommended that AZT be considered as a component to a treatment regimen in pregnancy. In cases where AZT cannot be used due to intolerance or resistance, a regimen that maximally suppresses viral load and that contains one or more drugs with good transplacental passage should be chosen.

Q: The document does not mention resistance testing. Is there anything unique about pregnancy that would differ from the standard recommendations for adults in general?

A: As with use of antiretroviral therapy, indications for testing for antiretroviral drug resistance in HIV-infected pregnant women is the same as in non-pregnant individuals. These indications include acute infection, virologic failure on an antiretroviral regimen, or suboptimal viral load suppression after initiation of antiretroviral therapy.

Data are conflicting regarding the influence of AZT resistance on perinatal transmission. Cases in which antiretroviral resistant virus was transmitted from mother-to-child have been reported. However, there are conflicting data on whether resistant virus is associated with increased risk of transmission despite prophylaxis. The presence of resistant virus is not necessarily associated with failure to prevent transmission. In PACTG 076, AZT resistance was rare and did not correlate with transmission, and in a recent study from the Pediatric AIDS Collaborative Transmission Study the rate of perinatal transmission did not differ between those with and without AZT resistance mutations. Further studies are needed to determine the best strategy to prevent perinatal transmission in the presence of AZT resistance. In circumstances where optimal maternal therapy precludes administration of AZT, it is recommended that intrapartum AZT and the 6-week infant AZT prophylaxis regimen should still be administered.

Q: This revised recommendation as well as the original one suggests "delaying initiation of therapy until after 10-12 weeks gestation." Is there any evidence that antiretroviral agents are dangerous during the first trimester or does this continue to represent a theoretical risk only?

A:Data are insufficient to determine whether use of antiretroviral drugs during the first trimester is safe. The available data from the Antiretroviral Pregnancy Registry suggest that use of AZT monotherapy during pregnancy is not associated with higher rates of congenital abnormalities than are observed in the general population. However, the number of outcomes among women receiving AZT during the first trimester is very small. The reports of first-trimester use of other drugs or drug combinations are very few in number, and preclude the ability to draw any conclusions. Since the first 10 weeks of pregnancy are the time of maximal organogenesis, drug exposure during this high-risk period should be minimized to the extent that it is possible.

Q:The recommendation is for the O76 regimen for women who refuse HAART or do not have criteria for HAART on the basis of guidelines for adults. Couldn't there be an argument that the goal of therapy for preventing perinatal transmission is to reduce the viral load to <1000 copies/ml regardless of the criteria of the DHHS guidelines for adults? Thus, a viral load of 8000, and CD4 cell count of 600, in a treatment naive woman would be treated with AZT monotherapy by current guidelines, but many would say that this is unlikely to bring optimal viral suppression and also encourage resistance?

A: Decisions regarding use of antiretroviral therapy during pregnancy should be made by the woman following discussion with her provider of the known and unknown benefits and risks of therapy. Women should be counseled that potent combination antiretroviral regimens have substantial benefit for their own health, and standard combination antiretroviral therapy is recommended as initial therapy for HIV-infected pregnant women whose clinical, immunologic or virologic status suggests the need for treatment if non-pregnant. Women should also be counseled that potent combination antiretroviral regimens may provide enhanced protection against perinatal transmission. Multiple studies have now confirmed that women with low or undetectable HIV RNA levels (e.g., <1,000 copies/ml) have extremely low rates of perinatal transmission (likely 2% or less), particularly when antiretroviral therapy has been received. Although the adult antiretroviral treatment guidelines have been modified to recommend initiation of therapy at higher viral load levels than previously (e.g., >30,000 copies/ml), use of potent combination HAART, including AZT whenever feasible, is recommended for pregnant infected women with HIV RNA >1,000 copies/ml. However, it is also important to counsel that there is no threshold below which lack of transmission can be assured. The risks of using combination therapy in women in whom such treatment would not ordinarily be recommended include the unknown long-term consequences for the infant who is exposed to multiple drugs during pregnancy and the current lack of data on pharmacokinetics for many antiretroviral drugs in pregnant women.

Use of antiretroviral prophylaxis is important even in women with HIV RNA <1,000 copies/ml. A meta-analysis of factors associated with perinatal transmission among 1,202 women with HIV RNA levels <1,000 at or near delivery found that the risk of transmission was only 1% among 834 women with RNA <1,000 who received antiretroviral therapy (primarily AZT alone) compared to 10% among 368 women with RNA <1,000 who did not receive therapy (Ioannidis JP et al. J Infect Dis 2001;183:539-45). The available data indicate that transient use of AZT monotherapy for prophylaxis for this subset of women is not harmful to their health. Time-limited use of AZT (started after the first trimester through labor) women with RNA <1,000 copies/ml is unlikely to be associated with the development of resistance given the limited viral replication and time-limited exposure to AZT. Data from PACTG 076, in which the women had generally low viral load and high CD4 count (i.e., those for whom therapy would generally not be recommended for their own health), indicated that transient AZT prophylaxis was not associated with high rates of development of AZT resistance. Additionally, the women who received AZT did not have increased long-term risk of disease progression compared with placebo women. PACTG 288 was a protocol that followed women who had been enrolled in PACTG 076 for a mean of 4 years postnatally. In this study, women who received AZT had similar rates of disease progression, CD4 cell decline, mortality, and development of resistant virus as those who received placebo. Thus, the current data do not imply that use of AZT prophylaxis is harmful to woman with low viral load and high CD4 cell count in terms of development of resistance or disease progression, and thus AZT monotherapy remains a reasonable option for such women.

Q: The guidelines now mention cesarean section. However, the clinical scenarios for treatment and for cesarean section are separated. Will the Guideline Working group be developing clinical scenarios which consider treatment and method of delivery at the same time? When might that happen?

A: The Working Group felt that issues related to choice of antiretroviral regimens involve specific concerns unrelated to choice of mode of delivery, and need to be discussed in detail separately. However, the clinical scenarios related to cesarean delivery already incorporate issues related to treatment. The basic recommendation is that elective cesarean delivery is recommended for women with HIV RNA levels >1,000 copies/ml. Although data to definitively show that elective cesarean delivery provides benefit in women receiving HAART who have HIV RNA levels >1,000 copies/ml, most members of the Working Group would recommend operative delivery in that setting regardless of type of antiretroviral therapy. Data are insufficient to show benefit of elective cesarean delivery in women with HIV RNA levels <1,000 copies/ml, and therefore operative delivery is generally not recommended for these women, regardless of antiretroviral therapy status. The Mode of Delivery scenario 1 involves women who have not received antiretroviral therapy; scenario 2 involves women who are receiving HAART with an initial virologic response but are expected to have HIV RNA >1,000 copies/ml at delivery; scenario 3 involves women receiving HAART who have undetectable viral load; and scenario 4 involves management of women who have had onset of labor or membrane rupture.

Q: We keep hearing about the potential toxicity of indinavir with its affect on bilirubin and possible nephrolithiasis in the infant. Is there any experience to show that this is a legitimate concern?

A: Enrollment is not yet complete in PACTG 358, a phase I clinical trial of indinavir/AZT/3TC. However, preliminary data from a small number of mother-infant pairs has not shown excessive hyperbilirubinemia or renal stones in the neonates of mothers who received this combination.

Q: The data for O76 are impressive, but the viral load data are even more impressive. Shouldn't the emphasis shift to the achievement of no-detectable virus at the time of delivery as the primary effort for preventing perinatal transmission?

A: As discussed previously, the guidelines for pregnant women currently recommend use of HAART for all HIV-infected pregnant women with HIV RNA levels >1,000 copies/ml. However, there is no threshold below which there is no risk of transmission.

It is important to recognize that there are likely several mechanisms by which antiretroviral therapy prevents transmission; reduction in maternal viral load is likely an important, but not the only, mechanism. Data suggest use of potent combination antiretroviral therapy is associated with very low rates of perinatal transmission (i.e., <2%) (Blattner W. XIII International AIDS Conference, Durban, South Africa, July 2000, Abstract LbOr4). However, transmission has been reported even among women receiving HAART who have viral loads below quantification. Additionally, genital viral load is likely to be an important risk factor for intrapartum transmission, and while in general reduction in both compartments has been described with antiretroviral therapy, discrepancies between peripheral blood and genital tract viral load reduction with antiretroviral therapy have also been reported.

Another important component of protection, particularly against intrapartum transmission, includes pre-exposure prophylaxis of the infant, through passage of the drug across the placenta to the fetus prior to intensive exposure to HIV in maternal genital secretions during birth. Therefore, in addition to lowering viral load, the ability of at least one drug in a treatment regimen to cross the placenta and reach virucidal levels in the fetus is an important component of a prophylactic regimen. Therefore, a regimen that maximally suppresses viral load and that contains one or more drugs with good transplacental passage is likely necessary for optimal reduction in transmission.

Q: For women who present very late in pregnancy or at the time of delivery without prior HIV serology, many advocate the use of rapid tests such as the SUDS. There is nothing in the current guidelines dealing with the issue of rapid testing in obstetrics. Please comment.

A: The U.S. Public Health Service Task Force Recommendations for the Use of Antiretroviral Drugs in Pregnant Women Infected with HIV-1 for Maternal Health and for Reducing Perinatal HIV-1 Transmission in the United States deal with guidelines related to antiretroviral drug therapy during pregnancy. Issues related to prenatal HIV counseling and testing are dealt with in another set of guidelines, the U.S. Public Health Service Recommendations for Human Immunodeficiency Virus Counseling and Voluntary Testing for Pregnant Women, published in July 1995. These guidelines recommended universal voluntary HIV counseling and testing for all pregnant women in the United States, including women who present late in pregnancy or at the time of delivery who have no prior testing or for whom HIV test results are unknown. The counseling and testing guidelines are currently in the process of being revised, and will discuss use of the now available rapid HIV tests during labor. The clinical trials that have shown that intrapartum/postpartum antiretroviral drug regimens (the HIVNET 012 nevirapine and the PETRA AZT/3TC regimens) can significantly reduce perinatal transmission demonstrate that effective interventions are available for women who learning HIV status even as late as the time of delivery.

Q: Are there specific concerns about mitochondrial toxicity with use of nucleoside analogue reverse transcriptase inhibitor treatment for HIV-infected pregnant women?