Infant prophylaxis for postnatal transmission

Polly Clayden, HIV i-Base

Two studies looked at giving prophylaxis to breastfed infants of HIV-positive mothers, who were negative at birth.


In an oral presentation Taha E Taha showed data from the PEPI trial. PEPI is an open label, controlled phase III trial conducted in Blantyre, Malawi. [1]

In this trial all mothers received single dose NVP for PMTCT. The infants were randomised to one of three arms immediately after delivery:  Arm 1 single-dose NVP + 1 week AZT (control arm); Arm 2 the control regimen plus extended daily NVP (ExtNVP) for 14 weeks; Arm 3 the control regimen plus extended daily NVP+AZT (ExtNVP/AZT) for14 weeks.

The primary endpoint was HIV infection at 9 months in infants who were HIV-negative at birth. This analysis evaluated 3016 infants enrolled before August 7, 2007, HIV negative at birth, for whom information on HIV status were available (1003 control, 1016 ExtNVP, and 997 ExtNVP/AZT).

Mothers were counselled to exclusively breastfeed and wean at 6 months. Mothers’ CD4 counts were similar in all three arms: control, median 401 cells/mm3 (IQR 263-587), n=921; ExtNVP, median 379 cells/mm3 (IQR 245.0-570.5), n=924; ExtNVP/AZT, median 400.5 cells/mm3 (IQR 280.0-581.0), n=902.

There was a very high rate of breastfeeding in this study from birth to 6 months. The investigators reported a reduction across all arms between 6 and 9 months (from 91% to 32% in control, 90% to 27% in ExtNVP, and 90% to 29% in ExtNVP/AZT).

At 14 weeks there were substantial differences in probability of infant infection: 8.4% in control arm, 2.8% % in ExtNVP and 2.8% in ExtNVP/AZT. The differences continued to 9 months: 10.6% control, 5.2% ExtNVP and 6.4% ExtNVP/AZT. At 9 months the probability of death was 8.9%, 6.8% and 6.3% in the control, ExtNVP and ExtNVP/AZT arms respectively. And the probability of infection or death was 16.8%, 10.6% and 11.2% in the three arms respectively.

The protective efficacy in the extended arms vs the control arm was 67%/66% at 14 weeks in the ExtNVP ans ExtNVP/AZT arms declining to 51%/40% at 9 months (and further decreasing over time: 23%/24% at 24 months). In a proportional hazards model, the risk factors for infant infection were ExtNVP vs control HR 0.56 (95% CI 0.41-0.76), p=0.0003; ExtNVP/AZT vs control HR 0.65 (95% CI 0.48-0.88), p=0.006 and maternal CD4 count (decrease of 100 units) HR 1.27 (95% CI 1.99-1.36), p<0.0001. Most deaths were caused by gastroenteritis and pneumonia. There was no difference in grade 2 or higher adverse events across all prophylaxis arms.


The Six Week Extended dose Nevirapine (SWEN) study evaluated a similar strategy but of a shorter duration. In an oral presentation, Jayagowri Sastry reported findings from this study. [2]

SWEN is a group of three separate but coordinated, randomised controlled trials conducted in Ethiopia, India, and Uganda to evaluate whether daily NVP given to breastfed infants until 6 weeks of age can decrease HIV transmission through breastfeeding.This presentation was a combined analysis from the three trials.

HIV-positive women breastfeeding their infants were enrolled. There were two prophylaxis arms: Arm 1, single dose NVP to mothers and infants. Multivitamin placebo to the infants from day 8 to 42. Arm 2, the single-dose NVP regimen plus 5 mg NVP daily from day 8 to 42 to the infants (SWEN). This study looked at the risk of HIV infection and death at 6 weeks and 6 months of age in infants HIV-negative at birth.

Maternal baseline CD4 was 397 cells mm/3 in the single dose NVP arm and 394 cells/mm3 in the SVEN arm. This data was from a modified intent-to-treat analysis including 986 single-dose NVP infants and 901 SWEN infants (excluding infants lacking specimens and those with indeterminate or HIV-positive at birth).

The investigators found that 6 weeks of age, SWEN infants had a 46% lower risk of HIV infection than the infants in the single-dose NVP arm (2.5% vs 5.3%; RR 0.536, 95%CI 0.336 to 0.855; p=0.009). At 6 months of age, SWEN infants had a non-significant, 20% lower risk of infection than single-dose NVP infants (6.9% vs 9.0%; RR 0.800, 95%CI 0.584 to 1.096; p=0.164). At 6 months of the mortality risk for the SWEN arm vs single-dose NVP arm was 1.1% vs 3.6% (RR 0.471, 95%CI 0.256 to 0.867; p=0.016).

The combined risks of post-natal HIV transmission or death in the SWEN arm vs the single-dose NVP arm were 3.7% vs 6.8% (RR 0.583, 95%CI 0.391 to 0.870; p=0.008) at 6 weeks and 8.0% vs 11.6% (RR 0.729, 95%CI 0.549 to 0.967; p=0.028) at 6 months, respectively.

The estimated cumulative probability of death or HIV transmission were RR 0.58, p=0.008; RR 0.7, p=0.026 and RR 0.73, p=0.028 at 6 weeks, 14 weeks and 6 months respectively. Serious adverse events were similar in both arms.

Resistance in SWEN

Inevitably this strategy should cause concern about resistance in infants who become HIV-positive despite receiving nevirapine prophylaxis.

In a second oral presentation from SWEN, Anitha Moorthly showed data from the Indian study comparing NVP resistance in infants receiving SWEN vs single-dose NVP by timing of HIV-1 infection and receipt of maternal single-dose NVP. [3]

In this analysis, infant DNA PCR was performed at 48 hours, 1, 2, 4, 6, 10, and 14 weeks, and 6, 9, and 12 months of age. Timing of infection compared 4 groups:  in utero (positive by 48 hours; n=22); peripartum/early breastfeeding (positive at week 1 to 6; n=19); late breast-feeding (positive at week 10 to 14; n=18), and very late breast-fed (positive at >/=6 months; n=35).

Median maternal CD4 counts were 316 cells/mm3 (IQR 238,454) and 320 cells/mm3 (IQR 194,522) in the single dose NVP and SWEN arms respectively. 57% 0f the women received maternal single dose NVP. 83/89 (93%) infant plasma samples could be genotyped. 76 infants met the inclusion criteria (able to define timing of infection) with sample taken 28 days since HIV diagnosis.

The investigators found higher rates of NVP resistance in SWEN infants infected within 6 weeks of life but lower rates of NVP resistance in infants infected after 6 weeks.

Table 1: Timing of infection

Timing of infection S/D NVP n=51 EXTNVP n=25
Infected at = 6 weeks 38% 92%
Infected at > 6 weeks 18% 5%
Infected in utero 50% 88%
Infected 1-6 weeks (post-partum / early BF) 27% 100%

The primary mutations were Y181C across both arms. There was no difference in mutations whether or not the mother had received NVP.

In multivariate analysis, the investigators found infants infected very late (>14 weeks) were 89% less likely to have NVP resistance than those infected earlier (in utero, peripartum, early breastfeeding, late breastfeeding) and these infants had mainly wild type virus.

With infection <= 6 weeks as reference the adjusted odds ratios for NVP resistance during late or very late transmission (population or clonal level) were: OR 0.75 (95%CI 0.21-2.69), p=0.658 and OR 0.11 (95%CI 0.02-0.48, p=0.004 for late and very late transmission respectively.

Following this presentation Dr Jeff Stringer remarked that according to his “back of an envelope calculation” of 1000 HIV exposed babies 21 may be HIV-negative but 53 may have NVP resistance and asked whether this was “too high a price to pay.”

And a poster authored by Jessica Church and coworkers evaluated resistance in the Ugandan infants. [4]

In this study, samples were available from 49/69 (71%) infants with HIV infection by 6 weeks of age (24 in the single-dose NVP arm; 25 in the SWEN arm). Maternal CD4 cell count, infant viral load, and HIV subtypes were similar in the both arms.

At six weeks NVP resistance was detected using the ViroSeq assay in a greater number of infants in the SWEN arm compared to the SD NVP arm (21/25 84% vs 12/24,50%, p=0.01).  A higher percentage of infants in the SWEN arm also had at least one NVP resistance mutation detected using the more sensitive LigAmp assay (19/25,79% vs 7/24, 35%, p=0.004).

In the SWEN arm, NVP resistance was not associated with the number of NVP doses received or the HIV status at birth. Among infants with resistance detected at 6 weeks, only 1 of 6 infants in the single-dose NVP arm had NVP resistance detected by ViroSeq at 6 months. All 7 infants in the SWEN arm still had detectable NVP resistance at 6 months.

Phenotypic resistance results were available for 42/49 (85.7%) of infants evaluated at 6 weeks. There was a higher percentage of infants with phenotypic resistance in the SWEN arm than in the single-dose NVP arm (19/22, 86.3% vs 9 /20, 45%, p=0.005).

The Indian and Ugandan analyses were consistent in showing SWEN infants were more likely to have NVP resistance than those who received only single-dose NVP.


Isn’’t this all getting rather over evolved?


Unless otherwise stated, all references are from the Programme and Abstracts from 15 Conference on Retroviruses and Opportunisitc Infections, 2-6 February 2008, Boston, USA.

  1. Taha T, Thigpen M, Kumwenda N et al. Extended infant post-exposure prophylaxis with antiretroviral drugs significantly reduces postnatal HIV transmission: The PEPI-Malawi Study. Oral abstract 42LB.
  2. Sastry J and The Six Week Extended Dose Nevirapine (SWEN) Study Team. Extended-dose nevirapine to 6 weeks of age for infants in Ethiopia, India, and Uganda: A randomised trial for prevention of HIV transmission through breastfeeding. Oral abstract 43.
  3. Moorthy A, Gupta A, Sastry J et al. Timing of infection is critical for nevirapine resistance outcomes among breastfed subtype C HIV-1-infected infants exposed to extended vs single-dose nevirapine prophylaxis: The India SWEN Study. Oral abstract 44.
  4. Church J, Omer S, Guay L et al. Analysis of NVP resistance in Ugandan infants who were HIV-infected despite receiving single-dose nevirapine vs single-dose NVP plus up to 6 weeks of daily NVP to prevent HIV vertical transmission. Poster abstract 635b.

Links to other websites are current at date of posting but not maintained.