Do HIV+ Pregnant Women on ART Need Intravenous Zidovudine During Delivery?

alt

Intravenous infusion of zidovudine (AZT; Retrovir) during delivery has been a key component in efforts to prevent mother-to-child HIV transmission, but it may not offer significantly additional protection for women with undetectable viral load, researchers reported in the September 15, 2013, edition of Clinical Infectious Diseases.

Intravenous zidovudine during labor and delivery is effective in reducing mother-to-child transmission among HIV positive women with virological failure (viral load at or above 1000 copies/mL), even if they are on antiretroviral therapy (ART) during pregnancy, according to an analysis from the French Perinatal Cohort (ANRS-EPF).

However, for the majority of women (77%) with well-controlled viral load (at or under 400 copies/mL) at delivery and without obstetrical risk factors (premature rupture of membranes, preterm delivery, fever, or bleeding), IV zidovudine was not significantly associated with a lower risk of HIV transmission.

These findings, note the authors, support recent changes in French and U.S. guidelines to no longer recommend systematic IV zidovudine during labor and delivery when viral load is low and no obstetrical risk factors exist.

The availability of ART in resource-rich settings has dramatically reduced mother-to-child transmission rates to approximately 1%. However, the precise role ART plays during pregnancy, labor, and delivery is difficult to determine, as transmission can happen at any time throughout.

Studies before the era of ART showed that most cases of mother-to-child transmission happened during labor and delivery.  So, in 1994, U.S. and French researchers designing the pivotal trial ACTG076-ANRS024 recommended IV zidovudine during labor or caesarean delivery as a link between pre- and post-natal prophylaxis. IV zidovudine was then included in guidelines in all resource-rich countries as a component of preventing mother-to-child transmission.

With the advent of effective ART, however the benefits of IV zidovudine have been questioned.

Following previous findings from this cohort (1997-2004), before ART was recommended for all pregnant women living with HIV, revised guidelines suggested IV zidovudine may not be necessary for women on ART with well-controlled viral load near delivery (not defined in the 2010 French guidelines and defined in the July 2012 U.S. guidelines as under 400 copies/mL).

This study comprised over 11,000 women with HIV, on ART, and not breastfeeding, who delivered at 90 centers throughout France between January 1, 1997 and December 30, 2010.

The authors looked at the effect of IV zidovudine on mother-to-child transmission according to maternal characteristics, viral load, and obstetric conditions.

There were no specific HIV treatment and obstetric care recommendations for women included in the cohort. From 1994, antenatal prophylaxis consisted of zidovudine; from 1997, it included dual nucleoside reverse transcriptase inhibitor (NRTI) therapy with elective caesarean section based on a risk/benefit evaluation.

The recommendation regarding ART for pregnant women needing it for their own health was revised in 2004 to include all pregnant women living with HIV. From 2002, elective caesarean section for preventing mother-to-child transmission was restricted to women not on ART and with a viral load over 400 copies/mL near delivery. Up until July 2010, IV zidovudine and newborn prophylaxis were systematically recommended.

The mother-to-child transmission rate was compared between women getting IV zidovudine and those who did not, analyzed according to viral load at delivery (<400, 400 to 999, or >1000 copies/mL), and stratified according to mode of delivery, term of delivery (under or over 37 gestational weeks), and type of post-natal prophylaxis.

Of the 11,538 women included in the analysis, 95% (10,984) got IV zidovudine while 554 did not. A protease inhibitor-based ART regimen was the last regimen prescribed during pregnancy for the majority of women (60%), with 7% using a non-protease inhibitor-based regimen, 3% using 3 or more NRTIs, 18% using 2 NRTIs, 10% using NRTI monotherapy, and <1% using an integrase inhibitor or CCR5 inhibitor.

For the analysis, CD4 counts and percentages and plasma viral load were obtained closest to the time of delivery and not more than 7 days after delivery. The median CD4 count at delivery was 461 cells/mm3. Close to 80% of newborns were given monotherapy prophylaxis.

The overall mother-to-child transmission rate among women with virological failure not taking IV zidovudine was significantly higher than the rate for those who did, 7.5% vs 2.9%, respectively.

Among newborns given zidovudine monotherapy, the mother-to-child transmission rate was 10.2%. However, giving newborns zidovudine plus lamivudine (3TC; Epivir) or a protease inhibitor-based ART regimen eliminated any significant difference in transmission rates between mothers who got IV zidovudine and those who did not, 4.1% vs 4.8%, respectively.

The analysis shows that IV zidovudine did not have an effect on mother-to-child transmission from women with well-controlled viral load and no apparent childbirth risk factors, but remained effective in reducing transmission from women with virological failure at delivery, even among those on ART during pregnancy.

While the study’s main strength is the large number of patients, the relatively small number (5%) not getting IV zidovudine is its main limitation.

Women not getting IV zidovudine were more likely to be over age 35, to have given birth previously, to have delivered preterm and vaginally, and to have high viral load at delivery. The authors hypothesize this may be because they presented in advanced labor, too late to start IV zidovudine.

Preterm delivery is a risk factor for mother-to-child transmission. Vaginal delivery in the absence of ART or with zidovudine monotherapy alone is linked to higher transmission rates compared to elective cesarean section.

However, there was a significant 5-fold increased risk of mother-to-child transmission without IV zidovudine than with (9.5% vs 1.8%) among women with virological failure delivering at term by elective cesarean section.

No transmission occurred among women with viral load under 400 copies/mL or even under 1000 copies/mL who did not get IV zidovudine, compared to 0.6% and 0.9%, respectively, among those who got IV zidovudine.

This finding led the authors to comment that given the large size of the cohort and few infected infants, the study lacked the power to evaluate the safety of withholding IV zidovudine in cases of obstetrical risk factors. They propose, while awaiting more data, that IV zidovudine be continued in women, even with low viral load, in cases of complicated deliveries.

These results, consistent with the authors’ previous findings, allowed them to look at IV zidovudine with an increased power and within the context of systematic ART recommended for all pregnant women since 2004. The proportion of women on protease inhibitor-based regimens increased from 38% in 1997-2004 to 86% in 2005-2010.

Evaluation of HIV-uninfected infants in the first 6 months of life showed that IV zidovudine was not associated with increased short-term hematological toxicity or lactate levels.

9/22/13

Reference

N Briand, J Warszawski, L Mandelbrot, et al (ANRS-EPF CO1-CO11 Study Group). Is Intrapartum Intravenous Zidovudine for Prevention of Mother-to-Child Transmission Still Useful in the Combination Antiretroviral Therapy Era? Clinical Infectious Diseases 57(6):903-914. September 15, 2013.