Selected Perinatal Transmission Studies at Retrovirus

Studies from New York City and North Carolina reported that use of elective cesarean section resulted in reduced vertical transmission, and is cost-effective. These studies showed increased use of combination therapy  and elective C-section among HIV-infected pregnant women in NYC and North Carolina. As well, it appears as though prenatal care is increasing in North Carolina and New York City. Both studies also showed further reduction in perinatal transmission rates. The studies look at the use and effectiveness of combination therapy in preventing perinatal transmission. In the NYC study, 12% received no prenatal care. In NYC the transmission rate was 3% in women receiving AZT plus 1 or more antiretroviral drugs when the neonate received AZT, and 6% when AZT alone was administered prenatally, intrapartum, and neonatal. A study in Newark, NJ showed the rate of perinatal transmission was significantly higher among women without prenatal care than those with prenatal care (24% vs 10%; Bardeguez et al, UMDNJ-NJ Medical School, Newark). Apparently, prenatal care is crucial to reducing incidence of vertical transmission.  

Cesarean Delivery Effectiveness and Safety is Questioned

In an article in Journal of the American Medical Association (JAMA; 1999; 281;1946) the author  suggests that at that time there was "..insufficient evidence to support the use of cesarean delivery to prevent perinatal transmission in most circumstances". I urge you to read the entire article on NATAP web site (http://www.natap.org/527199.html). But he refers to the only randomized trial at that time which showed women randomized to elective cesarean delivery (n=188) in this trial had a significantly lower rate of vertical transmission (1.8% vs 10.5%) than those randomized to vaginal delivery. But, in those women receiving zidovudine prophylaxis (n=236), the reduction in transmission risk was smaller and not statistically significant. When the trial participants were stratified by intended mode of delivery, a transmission rate of 4.3% was observed among zidovudine-treated women (n=117) assigned to vaginal delivery compared with 0.8% in those (n=119) assigned to cesarean delivery. He suggested that using combination therapy in a pregnant woman may be more appealing and effective, except possibly when the women presents late in gestation who for whatever reason has not received antiretroviral therapy. In that situation, an alternate consideration would be short-term AZT. He also suggests that c-section may be useful for a women who is unable to adhere to complex dosing regimens of ART.

The authors also raise the concern that c-section carries a substantially increased risk of perioperative complications in HIV-infected women.

Resistance. The following 2 studies raise concern about transmitting resistance to a newborn when monotherapy or suboptimal therapy is used to prevent transmission from mother to infant. The transmission of resistance can reduce treatment options for the baby and affect disease progression. J Pitt and investigators in WITS (Women and Infants Transmission Study) reported at Retrovirus that maternal AZT use during pregnancy was significantly associated with resistance. As well, AZT use during pregnancy was significantly associated with HIV progression in infants. Maternal plasma HIV RNA level had a trend towards association as well. They also reported that the infant's 1-2 month peak plasma viral load was significantly associated with progression. They concluded that while use of AZT during pregnancy by women reduces perinatal transmission, AZT use during pregnancy and early infant peak RNA levels may impact progression in infants who acquire HIV despite maternal AZT use. They said a larger studies are needed to address this concern. See more details below.

Researchers from the HIVNET-006 Study Group looked at Ugandan women for resistance six weeks after they received a single dose of nevirapine given at the onset of labor for prevention of HIV perinatal transmission. HIV reverse transcriptase sequences were analyzed from plasma collected from 14 women who received nevirapine single dose. The K103 NNRTI resistance mutation was detected in the samples of 1/3 women who transmitted and 2/12 women who did not transmit. Two of the 3 women had a K/N mixture at 103. The study authors said this suggested recent selection of the resistant variant. Pre-dose samples were available from 2 of these 3 women, and both pre-dose samples lacked the K103N mutation. Other nevirapine resistance mutations were absent in all 14 women ( A98G, L100I, V106A, V108I, Y181C, Y188C, G190A). Due to its long half-life, nevirapine can remain in the blood for several days or more at decreasing blood levels. Its possible that resistance was present but undetectable in other women at 6 weeks postdose. Sample drawn at 1 week postdoes are being analyzed. The authors said selection of the K103N mutation was unexpected in the absence of prior or concurrent AZT therapy, and suggests that mutations arising during antiretroviral therapy may be different for subtype B vs non-B HIV. Despite the possible development of this resistance mutation, nevirapine may still be useful in the developing world for several reasons including a cost/benefit analysis. For the developed world, where NNRTIs are readily available this data raises questions.  

Association of Maternal ZDV Use during Pregnancy and Infant ZDV Genotypic Resistance with Rapid Disease Progression among Infants in the WITS.

Factors associated with ZDV genotypic drug resistance (RT mutations at codons 41, 67, 70, 210, 215, or 219) in infant isolates from birth through 6 mos., and the impact of resistance on rapid disease progression (CDC Category 3 CD4 cell decline or Clinical Category C disease within the first year) were evaluated in 57 HIV+ infants enrolled in the WITS through March 1994. 17 women were treated with ZDV for HIV symptoms; 40 received no ZDV. Overall, 19/57 (33%) of infants were rapid progressors. While maternal ZDV use during pregnancy was significantly associated with resistance in infants [OR= 5.0 (95% CI: 1.5,17.5)], maternal plasma HIV RNA level had a trend towards association as well [OR = 11.4 (0.9,142.9) per each 500,000 copies]. Infant ZDV use during the first 6 mos. was not associated with resistance (p=0.48). Univariate analysis identified maternal ZDV use during pregnancy [OR = 4.9 (1.5,16.6)] and infant 1-2 mo. peak plasma HIV RNA levels [OR = 1.2 (>1.0,1.43) per 500,000 copies] as significantly associated with progression. Maternal plasma HIV RNA at delivery [OR = 5.4 (0.6,51.8) per 500,000 copies] and genotypic resistance in infants [OR = 2.6 (0.8,8.9)] were not significant. Infant ZDV use during the first 6 mos. was not associated with progression (p=.66). In 2-parameter models, maternal ZDV use during pregnancy remained significantly associated with progression [OR = 4.6 (1.1,19.3)] when adjusted for infant peak plasma HIV RNA [OR = 1.13(<1.0,1.4) per 500,000 copies]. Infant genotypic resistance after adjustment for infant peak plasma HIV RNA was not significant [adjusted OR= 3.5 (0.8,15.4].

Duration of Ruptured Membranes and Vertical Transmission of HIV.  In this study, J Read reported for the International Perinatal HIV Group.  Data was collected and analyzed on 4721 mother-child pairs from 5 European and 10 North American studies. The study authors reported that the likelihood of vertical transmission increased linearly with increasing DROM (duration of ruptured membranes), after adjusting for mode of delivery, receipt of antiretroviral therapy, maternal disease stage, and infant birth rate. The risk increased incrementally (2%) for each hour. The study authors suggested that a diagnosis of HIV at the time of delivery may potentiate the effect of DROM.. Of note, the mode of delivery (VagD or NECS) did not modify the relationship between DROM and vertical transmission. With AIDS, the probability of vertical transmission increased in a curvilinear fashion from 8% to 31% with DROM of two hours and 24 hours, respectively (p<0.01), abstract 659. Very little increase in the transmission rate occurred with the duration of rupture of membranes in women without AIDS, maybe due to a lower viral load.

Shapiro and others reported an interim analysis of study (PACTG 367), which  looks at prenatal records  of HIV-infected women delivering between January 1998 and May 1999 at 32 sites participating in the Pediatric AIDS Clinical Trials Group. See details below. Women received during pregnancy  either no ART, unknown ART, AZT alone, AZT/3TC, or multiple drugs with or without a PI. The rates of preterm birth and VPB were 20%/4%, and the rates of low birth weight and VLBW were 20%/5%. Among women who received ART with a PI vs without a PI, they found no differences in rates of preterm birth, low birth weight, or VLBW; but the VPB rate was lower with PI vs without PI.

Bryson and others from UCLA, Glaxo Wellcome and Agouron reported on the safety,  pharmacokinetics (PK), and antiviral activity of using nelfinavir in combination with AZT/3TC in 10 HI- infected pregnant women and their infants. She reported that pregnancy had no effect on maternal PK and placental transfer of nelfinavir was poor. And, there were significant decreases in viral load and increases in CD4s. See details below.

From a study conducted in Durban, South Africa, T-helper cell responses to HIV-1 envelope peptides were detected in 33/86 (38%) cord blood samples from HIV-1-seropositive and 0/9 from seronegative women. See details below.

Prevention of Mother-to-Child Transmission of HIV by Elective Cesarean Section: Costs, Outcomes, and Cost-Effectiveness

M Halpern and others from the Pediatric, Adolescent, and Maternal AIDS Branch of the NICHD reported HIV-infected women delivering by elective cesarean section (ECS), cesarean section prior to labor and rupture of membranes, have a lower risk of transmitting human immunodeficiency virus type 1 (HIV) to their infants.  However, the costs and outcomes of ECS to prevent HIV vertical transmission in the U.S. have not been evaluated.

Using a decision analysis model, the costs and outcomes of ECS versus vaginal delivery were assessed among HIV-infected pregnant women receiving no antiretroviral therapy (ART), zidovudine (ZDV) prophylaxis, or combination ART (ZDV and 3TC). Costs for mode of delivery, ART, postpartum morbidity, and pediatric care were included. Model outcomes were cases of vertical transmission avoided, years of the child's life saved, and dollars saved by avoiding vertical transmission. Cost-effectiveness was determined as the cost per case avoided or year of life saved.

Among women receiving no ART, ECS resulted in fewer HIV cases and decreased costs compared to vaginal delivery; ECS saved over  $2 in future HIV treatment costs for every $1 spent on ECS. Regardless of mode of delivery, ART resulted in fewer cases of vertical transmission. However, among women receiving ZDV prophylaxis or combination ART, ECS further decreased vertical transmission and, although increasing costs, was highly cost-effective. The cost-effectiveness of ECS was  $1,131 per HIV case avoided ($17 per year of life saved) and  $112,693 per case avoided ($1,700 per year of life saved) for women receiving ZDV prophylaxis and combination ART, respectively.

ECS is a cost-saving or cost-effective intervention to prevent vertical transmission of HIV, independent of receipt of ART. Further research is needed on the effectiveness of combination ART in preventing vertical transmission and on maternal and pediatric adverse events from exposure to this therapy. However, based on the findings of this study, ECS is likely to remain cost-effective over a wide range of clinical and economic scenarios.

Increased Use of Maternal Combination Antiretroviral Therapy and Elective Cesarean Section to Reduce Perinatal HIV Transmission in North Carolina.

Susan Fiscus and others North Carolina reported on their study to determine recent trends in the use of combination antiretroviral therapy (ART) and elective C-section among HIV-infected pregnant women in NC and their effects on perinatal transmission.

This study is a retrospective analysis of 261 infants born to HIV+ women in NC between Jan 1, 1998 and Sept 30, 1999, who had known infection status, ART, and/or delivery history compared with 525 infants with known infection status and ART history born in 1994-1997.

The use of any ART exposure to the infant increased from 62% in 1994 to 99% for the first 9 months of 1999. Maternal combination ART increased from 2% in 1996 to 68% in 1999. Mothers used 15 different combination regimens in 1998 and 10 in 1999. AZT+3TC (49%, 1998; 27%, 1999) and AZT+3TC+NFV (30%, 1998; 49%, 1999) were the most common combination regimens. 87% of women took AZT+3TC with or without some other antiretroviral agent. Elective C-sections increased from 9% in Jan/June 1998 to 43% in 1999. In 1998 and 1999, 50% and 54%, respectively, of HIV-infected women were identified during prenatal testing. They reported seeing a decrease in the number of HIV-infected infants born in the state - an average of 12/year for 1994-1996 to about 4-6/year for 1997-1999. Since 1998 none of the 36 infants born via elective C-section with known infection status has been infected, but 23/36 (63%) also received combination ART. 8/9 infected infants in 1998-1999 received either no ART (n=3), abbreviated ZDV monotherapy (n=3), or had mothers who did not adhere to the ART regimen (n=2).

Trends to Reduce Perinatal HIV Transmission in New York City.

Peters and others from the NYC Dept. of Health and the CDC looked at the trends in perinatal HIV prevention in NYC, including zidovudine (ZDV), other antiretroviral (ARV) use and perinatal HIV transmission.

Data were abstracted from infant records on maternal ARVs, neonatal ZDV, infant HIV infection status, and delivery mode for all identified HIV exposed infants at 22 sites participating in pediatric HIV surveillance in NYC.

Data are available for 2946 infants born in 1994-99 (57% of all NYC HIV positive births, per NYS DOH survey of childbearing women). Prenatal care (PNC) data are available for 1946: 12% received no PNC. In women with PNC, ≥82% had a prenatal HIV diagnosis, prenatal ZDV use increased from 40% ('94) to 81% ('99), and prenatal ZDV with other ARVs increased from 2% ('96) to 45% ('99). Among all women, C-sections increased from 13% ('94) to 32% ('99) (in '99 25% were to prevent HIV transmission). Use of multiple ARVs and C-sections was most prevalent in '98-'99, too recently for HIV status to be known for most births.

As you can see in the table below the percentage infected by HIV after birth was decreased to 3% when there was prenatal AZT administered with 1 or more additional ART drugs and AZT administered to neonate. This compares to 23% HIV-infection rate when no ART drugs were used.  

HIV infection status by timing of ARV use, among 1080 infants born '96-'97

INF=infected, UNF=uninfected, IND= indeterminate. Age-adjusted odds ratio: OR   

The authors concluded most women in PNC receive ARVs. Women on multiple ARVs have a low transmission rate.

Antepartum Antiretroviral Therapy and Pregnancy Outcomes in 462 HIV-Infected Women in 1998-1999 (PACTG 367).

The purpose of this study was to describe rates of preterm birth (<37 [PB] and <32 [VPB] weeks gestation), low birth weight (<2500g [LBW] and <1500g [VLBW]), stillbirth, structural abnormalities, and perinatal transmission among HIV-infected women according to antiretroviral therapy (ART) received during pregnancy.

This is an interim analysis of ongoing retrospective abstraction of prenatal records of HIV-infected women delivering between January 1998 and May 1999 at 32 sites participating in the Pediatric AIDS Clinical Trials Group (PACTG).

Of 945 women with known pregnancy outcome at >20 weeks gestation, 13% received no/unknown ART, 19% ZDV only, 20% ZDV/3TC, 8% combination therapy without protease inhibitor (PI), and 40% combination with PI therapy during pregnancy. The proportion of women who received ART increased each trimester (31%, 78%, and 94% received ART in the first, second and third trimester, respectively). PB/VPB and LBW/VLBW rates were 20%/4% and 20%/5%, respectively. Among women who received ART with PI vs. without PI, no difference was observed in rates of PB (relative risk [RR], 95%CI: 0.7, 0.5-1.1), LBW (RR 0.8, 0.6-1.3), or VLBW(RR 0.5, 0.2-1.2); the VPB rate was lower with PI vs. without PI (RR 0.2, 0.05-0.8). One of 3 stillbirths, 1 of 4 major structural anomalies and 6 of 15 minor structural anomalies were seen in fetuses that were exposed to PI. Twelve of 347 infants (3.5%, 95%CI 1.8-6.0%) with known HIV-status were diagnosed as infected. Infection rates according to maternal ART: none/unknown: 26%; AZT: 7.8%; AZT/3TC: 1.1%; multi-agent/no PI: 3.4%; multi-agent with PI: 1.1%

The authors concluded antepartum PI use did not appear to be associated with increased rates of adverse pregnancy outcomes. Few transmissions occurred among women who received multi-agent ART. Continued careful monitoring of short- and long-term adverse effects and benefits of antepartum ART use is imperative.

PACTG 353 A Phase I Study of Safety, Pharmacokinetics, and Antiviral Activity of Combination Nelfinavir (NFV), ZDV, and 3TC in HIV-Infected Pregnant Women and Their Infants.

In order to assess safety, pharmacokinetics (pk) and antiviral activity of nelfinavir (NFV) in combination with AZT and 3TC in HIV infected pregnant women and their infants 10 HIV infected pregnant women (14-34 weeks gestation) were enrolled and received oral (po) NFV (750mgtid), AZT (200mgtid), 3TC (150 mg bid), during gestation and postpartum (pp) and po NFV, 3TC, and IV AZT in labor.  Infants received 6 wks of po NFV (10mg/kg tid), ZDV (2.6mg/kg tid) and 3TC (2mg/kg bid).  Viral load, CD4 count, NFV pk and clinical and lab toxicities were assessed.  8 of 10 women completed the study; 1 had a therapeutic abortion, 1 had a fetal demise due to obstetrical complications.  Women tolerated the Rx without significant toxicity. There was a significant decrease in maternal plasma HIV RNA levels from baseline to delivery and pp (median -1.8, and -2.4 log, p=0.01) and  6/8 had plasma viremia of <400 HIV RNA copies/ml at delivery.  CD4 count and % increased from baseline to delivery and pp (median CD4 cells/ml 220/ 308/ 415) (mean CD4% 20,23,30) rsp. (p= 0.005). 8 infants had a median gestation age of 38 weeks (range 35-40wks) and all were HIV uninfected.  One infant who did not receive study drugs died with clinical sepsis and pulmonary hypertension (not study related). Transient grade 2/3 decreases in Hemoglobin and/or neutrophils occurred in 7 infants, which resolved with discontinuation of AZT.  Maternal nelfinavir Pk during pregnancy and 6 weeks postpartum was similar. The median maternal nelfinavir plasma levels at delivery was 2 mg/ml. Cord blood levels were undetectable or low in all infants.  At one week of life, the median nelfinavir levels were low (pre-dose 0.74ug/ml, Cmax 1.23 ug/ml). The authors concluded that the combination of nelfinavir and AZT and 3TC was well tolerated in HIV infected pregnant women and their infants and resulted in significant decreases in maternal plasma viremia and increases in CD4 cells.  Pregnancy had no effect on maternal nelfinavir pk and placental transfer of NFV was poor. The initial dose of 10mg/kg nelfinavir used in the infant was inadequate and further study of a higher dose (40 mg/kg bid) is underway.

HIV-1-Specific T-Helper Cell Responses Detected at Birth: Protection against Intrapartum and Breast Feeding-Associated Transmission of HIV-1.

Acquired HIV-1-specific cell-mediated immune responses have been observed in exposed-uninfected individuals, and it has been inferred, but not demonstrated, that these responses constitute a part of natural protective immunity to HIV-1. This study aimed to test this inference directly in the natural exposure setting provided by maternal-infant HIV transmission in a predominantly breast-fed population. Cord blood samples from infants of HIV-1-seropositive women enrolled in a study in Durban, South Africa, and of a control group of seronegative women, were tested for in vitro reactivity to a cocktail of HIV-1 envelope peptides (env) using a bioassay measuring IL-2 production in a murine cell line. Infants were followed with repeat HIV-1 RNA tests for up to 18 months to establish the timing of HIV infection. T-helper cell responses to HIV-1 envelope peptides were detected in 33/86 (38%) cord blood samples from HIV-1-seropositive and 0/9 from seronegative women (p=0.03).. Excluding 9 infants infected before delivery (venous blood collected on the day of birth positive for HIV-1 RNA) and 2 infants lost to follow-up, 0/28 infants with responses to env had detectable HIV-1 RNA on subsequent tests up to 18 months, compared to 8/47 (17%) unresponsive infants (p=0.02). Early acquired cellular immune responses to HIV-1, detectable in over a third of uninfected newborns of HIV-1-infected women, appear to provide complete protection against subsequent HIV-1 transmission at delivery and post-natally through breast-feeding. Protective cell-mediated immune responses from in utero exposure may be a more important mechanism of newborn protection against viral infection than previously recognized.