CROI: Extended infant nevirapine prophylaxis reduces HIV transmission through breastfeeding

Giving nevirapine prophylaxis to infants born to HIV-positive mothers for six to 14 weeks can reduce the risk of HIV infection through breastfeeding by half, according to findings from randomised studies conducted in Malawi, Ethiopia, India and Uganda presented on Monday at the Fifteenth Conference on Retroviruses and Opportunistic Infections in Boston.

In resource-limited settings, up to 40% of infants born to HIV-infected mothers have become infected by the end of breastfeeding period (which may last up to 18 months). Given the benefits and/or practical need to breastfeed, a variety of interventions directed at mother, infant, or both have sought to find an effective manner in which to interrupt transmission.

The World Health Organization recommends that mothers should receive zidovudine (AZT) from 28 weeks of pregnancy (or as soon as possible thereafter); single dose nevirapine and AZT/3TCduring labour, and AZT/3TC for seven days after delivery, while infants should receive single dose nevirapine and AZT for one week after birth.

But there is still concern that this regimen will not protect infants during the breastfeeding period, and over the past four years several studies have been conducted to assess the value of a longer period of antiretroviral prophylaxis for the infant.

The PEPI-Malawi study

In Malawi, over 3,000 infants who were uninfected at birth were enrolled in an open label, controlled study of three interventions. Following birth, infants were randomised to one of three groups for a period of 14 weeks. The “control” group received a single dose of nevirapine (NVP) and one week of zidovudine (AZT). A second group of infants received that regimen and extended NVP for 14 weeks (ExtNVP). The third group of infants received the control regimen plus extended NVP+ZDV for 14 weeks (ExtNVP/ZDV).

The primary endpoint of the study was HIV infection at 9 months in infants uninfected at birth. Breastfeeding duration was similar across the three groups: high (around 90%) from birth to six months and then declining to around 20% by nine months.

At 14 weeks, 10.6% of infants in the control group had seroconverted versus 5.2% in the ExtNVP group and 6.4%% in the ExtNVP/AZT group.

At nine months, 13% of infants in the control group had seroconverted versus just over 7% in the ExtNVP group and nearly 9% in the ExtNVP/ZDV group. Infant mortality was just under 9% in the control group and roughly 6.5% in each of the extended ARV therapy groups.

Combined seroconversion and mortality was 17% for infants in the control group at nine months. There was no statistical difference in combined seroconversion and mortality in the two extended therapy groups, with 11% in the ExtNVP group and 12% in the ExtNVP/ZDV group.

Most deaths were caused by gastroenteritis and pneumonia. Concerning safety, the number of grade 2 or higher adverse events were similar in each of the treatment arms.

Use of NVP alone was just as effective as use of NVP/ZDV in reducing HIV transmission or mortality.

Similar results from combined clinical trial outcomes in Ethiopia, India, and Uganda

In similar, but shorter duration, clinical trials in Ethiopia, India, and Uganda, results were combined to evaluate outcomes in overall transmission risk at six months. The studies were testing the SWEN approach – Six Week Extended Nevirapine. All women received 200 mg of NVP during labour. All infants received NVP (2mg/kg) at delivery, and were then randomised to receive either NVP 5mg daily from day 8 to day 48, or an active control dose of liquid multivitamins. The study was blinded to mothers and investigators by using opaque dropper bottles or pre-prepared syringes to administer the nevirapine or active control.

At birth, 986 infants received sd-NVP (2 mg/kg) and 901 received the SWEN regimen. There were no significant differences between the study arms in baseline characteristics.

At six weeks of age, infants in the SWEN study had a 50% lower risk of seroconversion as measured by HIV DNA PCR than did infants who received sd-NVP (2.5% vs. 5.27%). At six months, infants in the SWEN study had a 20% lower risk of seroconversion than did infants who received sd-NVP (roughly 7% versus 9%), but this difference was not statistically significant.

The risk of mortality in the control arm at six months was 3.6% vs. 1.1% in the SWEN study. The combined risk of either seroconversion or death in the sd-NVP control arm was 11.6% vs. 8% in the extended NVP group and this difference was statistically significant. Serious adverse events were similar in both groups.

Again, in these results, extended infant prophylaxis with NVP was safe and resulted in greater reductions in both HIV-free survival and mortality in breastfed infants than did sd-NVP.

Nevirapine resistance in infants in the SWEN study

An analysis of nevirapine resistance in infants in the SWEN study was also carried out among Indian participants, since there is a concern that infants who become infected with nevirapine-resistant virus during delivery might not benefit from this approach, and that this might partially explain the transmission rate in those who received the SWEN regimen.

HIV-1 DNA polymerase chain reaction (PCR) was done within 48 hours postpartum, at weeks 1, 2, 4, 6, 10, and 14, and at month 6, 9, and 12. Population-based genotyping was done at a median 28 days after a positive DNA-PCR test.

The timing of HIV infection was divided into four periods:

• in utero, if HIV-positive by 48 hours (23 infants)
• HIV-positive at week 1 through 6 (19 infants)
• HIV-positive at 10 to 14 weeks (18 infants) and
• HIV-positive at 6 months or beyond (19 infants).

Of those infants who were HIV-diagnosed within 6 weeks of age, those who had received extended-dose NVP had greater NVP resistance (11/12) than did infants who received sd-NVP only at birth (12/30). The type and extent of resistance (>1 mutation) did not vary according to NVP dose with the most common mutations being Y181C, followed by K103N and Y188C.

NVP resistance in infants at 14 weeks was not affected by whether the mother had received sd-NVP intrapartum or not.

Practical implications

Although the findings suggest that the extended nevirapine regimen is safe and effective, post-presentation discussion revealed concern about the high rate at which infants, infected despite the extended nevirapine regimen, acquired nevirapine resistance. Dr Jeff Stringer of the Centre for Infectious Disease Research in Zambia pointed out that based on the findings of the SWEN study, if 1000 infants were treated with the regimen, approximately 21 extra infections might be averted compared to the use of single-dose nevirapine, but approximately 53 infants might acquire nevirapine resistance with unknown consequences for their future prognosis.

Speaking at a press conference to unveil the results, Dr Elaine Abrams of Columbia University’s ICAP programme, which has pioneered combined care of mothers and infants in resource-limited settings with PEPFAR funding, said: “These results are likely to have an enormous impact on the field and on WHO guidelines.”

But, she acknowledged, “part of the challenge of implementing these findings will be to link prevention of mother to child transmission services and maternal child health care.”

One of the attractions of the extended nevirapine regimen is that it is relatively safe, said Dr Taha Taha of Johns Hopkins University Bloomberg School of Public Health, but the risk of anaemia and neutropenia would need to be monitored, and facilities for infant diagnosis using HIV DNA-PCR would also be needed.

However, others were less enthused about the importance of the findings. Dr Francois Venter, chair of the South African HIV Clinicians Society, told aidsmap: "There is a trend for PMTCT intervention to become more and more complicated, when what I think we need to be working towards is HAART for mothers and starting treatment when the CD4 count falls below 350 cells."

The findings are likely to be debated in relation to findings from a study of the effects of maternal HAART on HIV transmission to breastfeeding infants, also presented during Monday's conference session (see report here)