8.1 Infant post-exposure prophylaxis
(See Table 1 for quick reference guides to infant ARV regimens and infant dosing.)
Table 1. Infant doses of ARV therapy (all treatment for 4 weeks except nevirapine)
|Zidovudine (ZDV, AZT)|| |
Term (>34 weeks):
4 mg/kg twice daily
Premature (30–34 weeks):
2 mg/kg twice daily for 2 weeks then 2 mg/kg three times a day for 2 weeks
Premature (<30 weeks):
2 mg/kg twice daily for 4 weeks
Term: 1.5 mg/kg four times a day
Prem: 1.5 mg/kg twice daily
Combo (+ lamivudine)
Moodley 2001 
Boucher 1993 
Capparelli 2003 
Boucher 1993 
Frasca 2009 
Anaemia, neutropenia – more common with combination therapy in mother and infant.
In French study of ZDV + lamivudine a small proportion of infants required either blood transfusions or early stop of therapy.
Transient lactic acidaemia has been observed in HIV-uninfected infants exposed to HAART in utero and/or ZDV neonatally 
|Lamivudine (3TC)||2 mg/kg twice daily|| |
Combo (all with ZDV)
Combo (+ nelfinavir)
Mandelbrot 2001 
Moodley 2003 
Durand-Gasselin 2008 
Hirt 2011 
Mirochnick 2011 
|Anaemia, neutropenia (but less common than with ZDV). More common with combination therapy in mother and infant.|
|Abacavir (ABC)||2 mg/kg twice daily||Mono||Jullien 2005  ||Hypersensitivity reaction has not been noted in infants (only small numbers treated)|
|Didanosine (ddI)||60 mg/m2 twice daily||Mono||Wang 1999  ||Much better absorbed on an empty stomach. Difficult to separate dosing from feeding. May cause gastrointestinal symptoms. Variable absorption in neonates|
|Emtricitabine (FTC)||2 mg/kg as a single dose (with 13 mg/kg of TDF) within 12 h after birth||Combo (with TDF)||Hirt 2011  || |
Mothers received two tablets of TDF/FTC at onset of labour and then one tablet daily for 7 days postpartum.
This dose resulted in high FTC levels in neonates.
Can cause neutropenia, anaemia
|1 mg/kg as a single dose immediately after birth.||Combo (with ZDV and NVP)||Hirt 2009  ||Dose based on pharmacokinetic modelling study|
|Tenofovir (TDF)|| |
13 mg/kg as a single dose within 12 h of life.
On the first day of life, neonates received a single dose of NVP syrup (2 mg/kg), within the 12 h after birth a single dose of TDF oral solution (13 mg/kg) and a single dose of FTC oral solution (2 mg/kg), and for 7 days ZDV syrup (4 mg/kg every 12 h).
|Combo (with NVP, FTC and ZDV)||Hirt 2011  || |
Single dose administered to neonate after the mothers had received two tablets of TDF/FTC at delivery.
Associated with renal dysfunction: monitor renal function in neonates.
|Nevirapine (NVP, NEV)|| |
Daily dosing regimen:
2 mg/kg once a day for 1st week then 4 mg/kg once a day for 2nd week then stop.
Use 4 mg/kg once a day for 2 weeks if mother has received more than 3 days NVP.
one 2 mg/kg dose 48–72 h from birth
|Shetty 2004  ||Daily dosing regimen from HIVNET 023 (breastfeeding prophylaxis study)|
|2 mg/kg as a single dose on the first day of life plus ZDV 4 mg/kg every 12 h for 7 days.||Combo (with ZDV)||Hirt 2009  ||Mothers received ZDV 300 mg twice a day to the delivery date, one tablet of NVP (200 mg) and two tablets of TDF/FTC at start of labour, and one tablet of TDF/FTC daily for 7 days postpartum.|
|Nelfinavir (NEL/NFV)||50–75 mg/kg twice daily||Combo (with ZDV+ 3TC)|| |
Mirochnick 2011 
|Nelfinavir 250 mg tablets can be dispersed in water.|
|Lopinavir/ ritonavir|| |
300 mg/m2 twice daily
1–2 kg: 40 mg every 12 h
2–6 kg: 80 mg every 12 h
Jullien 2006 
Verweel 2007 
Chadwick 2008 
Chadwick 2011 
Urien 2011 
Some pharmacokinetic studies have suggested that a twice-daily dose may give low levels in neonates. Frequent dose adjustment for weight gain is advisable.
Adrenal dysfunction reported in newborns. Monitor electrolytes. Avoid in premature babies . FDA recommendation (August 2011): the use of Kaletra oral solution should be avoided in premature babies until 14 days after their due date, or in full-term babies <14 days of age unless a healthcare professional believes that the benefit of using Kaletra oral solution to treat HIV infection immediately after birth outweighs the potential risks. In such cases, FDA strongly recommends monitoring for increases in serum osmolality, serum creatinine and other signs of toxicity.
900 mg/m2 once daily Mon/Wed/Fri
<6 months: 120 mg once daily Mon/Wed/Fri
6–12 months: 240 mg once daily Mon/Wed/Fri
|PCP prophylaxis||Simmonds 1995  ||May cause rash, bone marrow suppression. Give to infants aged over 4 weeks born to mothers with a higher risk of transmission|
8.1.1 Zidovudine monotherapy is recommended if maternal VL is <50 HIV RNA copies/mL at 36 weeks' gestation or thereafter before delivery (or mother delivered by PLCS while on zidovudine monotherapy). Grading: 1C
For women with fully suppressed HIV and a history of zidovudine resistance see discussion below.
Zidovudine monotherapy for the infant has been part of the PMTCT strategy since publication of the ACTG 076 results . The relative contributions of the antenatal, peripartum and infant components have been difficult to quantify. In ACTG 076 neonatal zidovudine 2 mg/kg every 4 h (five doses) was given for 6 weeks.
Monotherapy for the infant is appropriate when there is a very low risk of HIV transmission. This occurs when a mother on combination therapy delivers with a VL <50 HIV RNA copies/mL. The neonate should receive single-drug therapy for 4 weeks; this is practically easier for the family and reduces the risk of adverse events. With many years of experience, twice-daily zidovudine monotherapy is the neonatal treatment of choice, whatever the maternal ART combination.
For infants born to mothers on fully suppressive ART, zidovudine monotherapy PEP remains reasonable even where the mother has a previous history of zidovudine exposure with resistance (thymidine-associated mutations). On HAART, the risk of transmission in the mother with fully suppressed viral replication is extremely low ( about 0.1%), and although history of zidovudine resistance in maternal virus and infant PEP regimen has not been dissected, the frequency of transmission of zidovudine-resistant virus is concomitantly very low. Data from the era when only maternal zidovudine monotherapy was available indicate preferential transmission of wild-type over zidovudine-resistant virus when a mixed population of virions are present . In the Swiss cohort, none of six infants born to mothers harbouring zidovudine-resistant HIV (based on codon 215 analysis only) became infected . In a subset of participants of the ACTG 076 study, the prevalence of low-level zidovudine resistance was 4.3% (mutation at codon 70) and no significant increase in the risk of transmission was observed after adjusting for VL at delivery (OR 4.8; with wide 95% CI 0.2–131; P = 0.35) . High-level resistance was not reported and the median CD4 cell count in the women was 540 cells/μL. In retrospective cohort studies from France  and the USA , 20% and 8.3%, respectively, of HIV-positive newborns had zidovudine-resistance mutations after maternal zidovudine prophylaxis. In the WITS, lower CD4 cell count and higher HIV VL at delivery were associated with increased risk of transmission while in the multivariate analysis, the presence of at least one mutation associated with zidovudine resistance was also associated with an increased risk of transmission (OR 5.15; 95% CI 1.4–18.97) . With infant feeding patterns, it is difficult to separate drug dosing from feeds, so drugs without food restrictions are preferred, an advantage of zidovudine. Important in this age group, where therapeutic options are more limited than in older children and adults, should transmission occur multidrug resistance is avoided. However, some clinicians prefer to choose another ARV, with no history of maternal resistance, for infant post-exposure monotherapy. The established alternatives, nevirapine and lamivudine, have potent ARV effect but a low (single-point mutation) barrier to resistance. The dosing and safety issues with newer therapies, such as lopinavir/ritonavir, are outlined below. It is therefore suggested that neonatal zidovudine monotherapy remains a reasonable approach for infants born to mothers with a HIV VL <50 HIV RNA copies/mL plasma, even if there is a history of zidovudine resistance. Further investigation of the national cohort data to address this question is under way.
Where a low transmission-risk mother (see Section 5: Use of antiretroviral therapy in pregnancy) chooses zidovudine monotherapy plus PLCS, the infant should receive zidovudine monotherapy .
There are two situations where triple combination PEP for neonates is advised:
- Post-delivery infant-only prophylaxis: mother found to be HIV positive after delivery, which is only effective if given within 48–72 h of birth.
- Detectable maternal viraemia (>50 HIV RNA copies/mL) at delivery, mother may be on HAART or not:
- delivery before complete viral suppression is achieved (e.g. starting HAART late or delivery premature);
- viral rebound with or without resistance, with or without poor adherence;
- unplanned delivery ( e.g. premature delivery before starting ART or late presentation when maternal HIV parameters may be unknown).
8.1.2 Infants <72 h old, born to untreated HIV-positive mothers, should immediately initiate three-drug ART for 4 weeks. Grading: 1C
There is one large RCT of combination therapy in neonates born to mothers who did not receive any ART before delivery (n = 1684, in Brazil, Argentina, South Africa and the USA) . Infants were randomly allocated at <48 h of age to: 6 weeks of zidovudine monotherapy; or 6 weeks of zidovudine with three doses of nevirapine in the first week of life; or 6 weeks of zidovudine, with nelfinavir and lamivudine for 2 weeks. Overall, in this high-risk group, the HIV transmission rate was 8.5%, and in multivariate analysis, only ART arm and maternal VL were significantly associated with transmission. For infants uninfected at birth, transmission was twofold higher in the zidovudine-alone arm compared to the multiple ART arms (P = 0.034). There was no significant difference in transmission rates between the two multiple ARV arms and neonatal neutropenia was significantly higher in the three-drug arm.
In a randomized African study, babies born to mothers presenting at delivery received single-dose nevirapine or single-dose nevirapine and 1 week of zidovudine. Of those HIV negative at birth, 34 (7.7%) who received nevirapine plus zidovudine and 51 (12.1%) who received nevirapine alone were infected (P = 0.03): a protective efficacy of 36% for the dual combination . However, in two other randomized African studies where the mothers received short-course ART, for infants uninfected at birth there was no significant difference in transmission rate at 6 weeks for dual vs. monotherapy short-course regimens to the infant: zidovudine plus lamivudine vs. nevirapine ; or zidovudine plus nevirapine vs. nevirapine .
PEP for the infant of an untreated mother should be given as soon as possible after delivery. There are no studies of time of initiation of combination PEP, but in a US cohort study a significantly reduced risk of transmission was only observed in infants commenced on zidovudine when this was started within 48 h of birth . For this reason, infant PEP should only be started where a mother is found to be HIV positive after delivery if it is within 48–72 h of birth.
NSHPC data from the UK and Ireland 2001–2008 demonstrate how the clinical practice of combination PEP in neonates has increased over time . In total, 99% of 8205 infants received any PEP, and for the 86% with data on type of PEP, 3% received dual and 11% triple. The use of triple PEP increased significantly over this period, from 43% to 71% for infants born to untreated women, and from 13% to 32% where mothers were viraemic despite HAART. HIV infection status was known for 89% of infants with information on PEP; 14.7% of infants who received no PEP were infected (five of 34, all born vaginally to untreated mothers), compared to 1% of those who received any PEP (72 of 7286). Among infants born vaginally to untreated mothers, those who received PEP were significantly less likely to be infected than those who did not [8.5% (four of 47) vs. 45.5% (five of 11), P = 0.002]. However, in this cohort study, because of the overall low rate of transmission and selective use of triple PEP for infants at higher risk of HIV, it was not possible to explore the association between type of PEP and infection status.
8.1.3. Three-drug infant therapy is recommended for all circumstances other than Recommendation 8.1.1 where maternal VL at 36 weeks' gestation/delivery is not <50 HIV RNA copies/mL. Grading: 2C
Delivery with a detectable maternal VL (>50 HIV RNA copies/mL) is not uncommon. The virus may never have been suppressed due to: premature delivery; poor adherence; very high starting maternal VL (>100 000 HIV RNA copies/mL); or late commencement of HAART; or there may have been viral rebound during gestation due to poor adherence or development of resistance.
There are no randomized trials of combination therapy PEP for infants where mothers are receiving HAART. In a French study, transmission rates with dual therapy (zidovudine and lamivudine) to both the neonate and mother (1.6%) were lower than zidovudine monotherapy reported in historical controls (6.8%; OR 0.22; 95% CI 0.2–0.5) .
The strength of recommendation is proportionate to the estimated risk of transmission. Thus, benefit of additional neonatal therapy is anticipated at higher VLs, in circumstances where resistance is suspected or confirmed and where VL is increasing despite treatment. As with the recommendations regarding PLCS at VLs <400 HIV RNA copies/mL, favourable trends can be considered in the risk assessment. Despite the lack of evidence for its use, NSHPC data indicate a trend towards increasing use of triple-neonatal PEP.
When an infant has been started on triple-combination PEP because the maternal VL is >50 HIV RNA copies/mL at 36 weeks and subsequently a delivery maternal VL is <50 HIV RNA copies/mL, then it is reasonable to simplify the infant PEP to monotherapy.
Choice of triple combination post-exposure prophylaxis for neonates
Most neonates born in the UK to mothers known to have HIV will be exposed to ART in utero, during delivery and after birth for the first 4 weeks of life. The range of cARTs to which neonates are being exposed in utero continues to increase. Neonatal drug metabolism is generally slower than that of older infants or children and premature neonates have even less efficient metabolism. Owing to a lack of neonatal pharmacokinetic and efficacy studies and suitable formulations, ART dosing regimens remain restricted to a small proportion of the ARV drugs currently manufactured (Table 1). Small pharmacokinetic studies have been performed (zidovudine , lamivudine ,, tenofovir , emtricitabine ) and dosing regimens are available for most of the nucleoside analogues and for abacavir from age 1 month , while limited study of didanosine in neonates suggests that the pharmacokinetics are highly variable . The pharmacokinetics of nevirapine in neonates has been described in more detail ,,[268-270]. Pharmacokinetic-supported dosing is available for the PIs nelfinavir  and ritonavir-boosted lopinavir (based on HIV-1 infected infants initiating therapy in the first 6 weeks of life) [271-273] and a study that included some infants treated from birth . However, evidence of adrenal suppression has been documented in some neonates treated with lopinavir/ritonavir, particularly when preterm , in addition to case reports of cardiac, renal and neurological toxicity, especially in, but not restricted to, premature infants, and including one death during PEP with lopinavir/ritonavir . No effects have been observed with maternal lopinavir/ritonavir in the absence of neonatal dosing. It remains unclear whether these effects are related to lopinavir/ritonavir specifically or could be seen with other ritonavir-boosted PIs. The Writing Group therefore recommends that this PI should be avoided in routine infant PEP and should only be prescribed to preterm neonates in exceptional circumstances. Its use should only be considered after seeking expert advice and where there is multidrug resistance. Close metabolic monitoring in hospital should be undertaken. Nelfinavir, the only other PI with an infant-dosing regimen, will be withdrawn in the near future and will no longer be available for prescription in the UK or elsewhere in Europe. See the CHIVA website for dosing updates (http://www.chiva.org.uk).
In contrast to the PIs, nevirapine efficiently crosses the placenta (see below) and is well absorbed by the neonate . Neonatal metabolism of nevirapine is induced where there has been antenatal in utero exposure ,; if this drug is given to the neonate when the mother has taken it for 3 or more days, the full dose of 4 mg/kg per day should be started at birth, rather than the induction dose of 2 mg/kg per day (Table 1). Owing to its long half-life, nevirapine should be stopped 2 weeks before co-prescribed ARV drugs with shorter half-lives to reduce the risk of nevirapine monotherapy exposure and the development of NNRTI resistance should transmission have occurred.
The only licensed ART available for intravenous use in sick and/or premature neonates, unable to take oral medication, is zidovudine ,. Reduced oral and intravenous dosing schedules for premature infants are available (Table 1). The fusion inhibitor, enfuvirtide does not cross the placenta. Although intravenous enfuvirtide (T20) has been given to a small number of infants born to mothers with multidrug resistant HIV, no formal neonatal pharmacokinetic studies for enfuvirtide have been conducted to date. The dose used has been adapted from a paediatric subcutaneous treatment study  and an adult intravenous dosing study .
For infants born to ART-naïve women or where drug resistance is unlikely, zidovudine, lamivudine and nevirapine is the well-tolerated combination therapy regimen with most experience (see Table 1 for dosing). Infants born to non-naïve mothers, or mothers known to have ART resistance, may require other combinations (seek expert advice).
Resistance testing should be carried out in the mother. Where this is not available, choice of treatment has to be made based on history of drug exposure and any previous resistance data in the mother. If the infant is infected, then the first HIV-positive sample should also be tested for the resistance pattern of the transmitted virus.
Intravenous antiretroviral therapy in the neonate.
The very premature neonate is at risk of necrotizing enterocolitis if enteral feeding is commenced too soon or increased too rapidly. It is not known whether very early enteral administration of ART can exacerbate this risk. In a large French case-controlled study of cases of necrotizing enterocolitis, being an infant of a mother with HIV was associated with an increased risk of necrotizing enterocolitis (OR 6.63; 95% CI 1.26–34.8; P = 0.025), although the numbers were too small to ascertain the effect of maternal and/or infant ART . Premature infants should be commenced on intravenous zidovudine, but once enteral feeding is established, zidovudine may be given enterally and the premature dosing regimen should be used (Table 1). Enfuvirtide is the only other ARV administered parenterally, usually subcutaneously, in adults and children. An unlicensed intravenous dosing regimen has been adapted for use as part of cART in neonates at risk of multiresistant HIV (seek expert advice) .
8.1.4 Neonatal PEP should be commenced very soon after birth, certainly within 4 h. Grading: 1C
There are no clear data on how late infant PEP can be initiated and still have an effect, but all effective studies of infant PEP have started treatment early and animal data show a clear relationship between time of initiation and effectiveness [282-284]. Immediate administration of PEP is especially important where the mother has not received any ART.
8.1.5 Neonatal PEP should be given for 4 weeks. Grading: 1C
In the original ACTG 076 study, zidovudine was administered for 6 weeks after birth and this subsequently became standard of care . Simplification to zidovudine twice daily for 4 weeks has become common practice in the UK and data from the NSHPC suggest that regimens adopting this strategy remain highly effective . Recent cohort studies from Ireland  and Spain  have demonstrated efficacy and reduced haematological side effects with 4 vs. 6 weeks of neonatal zidovudine. In a Thai study, where a short course of 3 days of neonatal monotherapy zidovudine PEP was compared with 6 weeks, there was no significantly increased HIV transmission where the mother received zidovudine monotherapy from 28 weeks' gestation . Whether 4 weeks of zidovudine is necessary for infants born to mothers on HAART with fully suppressed HIV is not known, shorter courses may be considered in the future.
8.2 Pneumocystis pneumonia prophylaxis
8.2.1 PCP prophylaxis, with co-trimoxazole, should be initiated from age 4 weeks in:
- All HIV-positive infants. Grading: 1C
- In infants with an initial positive HIV DNA/RNA test result (and continued until HIV infection has been excluded). Grading: 1C
- Infants whose mother's VL at 36 weeks' gestational age or at delivery is >1000 HIV RNA copies/mL despite HAART or unknown (and continued until HIV infection has been excluded). Grading: 2D
Primary PCP in infants with HIV remains a disease with a high mortality and morbidity. However, as the risk of neonatal HIV infection has fallen to <1% where mothers have taken up interventions, the necessity for PCP prophylaxis has declined and in most European countries it is no longer prescribed routinely. However, co-trimoxazole, as PCP prophylaxis, should still be prescribed for infants born to viraemic mothers at high risk of transmission. The infant's birth HIV molecular diagnostic test (see below) and maternal delivery VL should be reviewed before the infant is aged 3 weeks. If the HIV molecular diagnostic test taken in the first 24 h is positive, the infant should be reviewed before 4 weeks for an early repeat test and to be started on co-trimoxazole prophylaxis, which should be continued if the HIV infection is confirmed, and stopped if infection is excluded (see section on diagnosis below). Infants with a first positive HIV molecular diagnostic test at age 6 or 12 weeks should be started on co-trimoxazole prophylaxis until HIV infection is confirmed or excluded (see Table 1 for dose).
If the birth HIV diagnostic test is negative, and the maternal delivery VL is <1000 HIV RNA copies/mL, there is no need to start co-trimoxazole prophylaxis and the baby can be seen routinely for a second HIV diagnostic test at age 6 weeks.
Co-trimoxazole prophylaxis against PCP is effective, but there are no data on when to initiate it in infants of indeterminate HIV status being followed up after in utero exposure to HIV. A maternal VL of 1000 HIV RNA copies/mL is an arbitrary cut-off to define infants at higher risk of transmission, in whom it is recommended to start prophylaxis until lack of transmission has been established.
8.3.1 Infants born to HIV-positive mothers should follow the routine national primary immunization schedule. Grading: 1D
Generally, BCG vaccine should only be given when the exclusively formula-fed infant is confirmed HIV uninfected at 12–14 weeks. However, infants considered at low risk of HIV transmission (maternal VL <50 HIV RNA copies/mL at or after 36 weeks' gestation) but with a high risk of tuberculosis exposure may be given BCG at birth.
Where the mother is coinfected with HBV, immunization against HBV infection should be as per the Green Book and does not differ from management of the HIV-unexposed infant .
With sensitivity to concerns about confidentiality, families should be strongly encouraged to inform primary health carers, including midwives, health visitors and family doctors about maternal HIV and indeterminate infants. This will enable the local team to give appropriate support and advice, especially regarding infant feeding and where the infant or mother is unwell.
8.4 Infant feeding
8.4.1 All mothers known to be HIV positive, regardless of ART, and infant PEP, should be advised to exclusively formula feed from birth. Grading: 1A
It is well established that HIV can be transmitted from mother to child by breastfeeding [289-291]. RCT evidence from Kenya puts the transmission rate at 16% over 2 years, accounting for almost half the total MTCTs . Complete avoidance of breastfeeding removes this risk altogether [291-293] and is the current standard of care in the UK ,. This is in line with previous World Health Organization (WHO) guidance, that exclusive feeding with infant formula milk should be recommended for women with HIV where it is affordable, feasible, acceptable, sustainable and safe .
Recently, cohort [296-299]; and RCT ,, data from Africa have shown that ART can significantly reduce the risk of HIV transmission from breastfeeding. This is in settings where breastfeeding is not affordable, feasible, acceptable, sustainable and safe, and mortality from formula feeding outweighs additional mortality from HIV transmission by breastfeeding ,. WHO guidance remains that in countries where formula feeding is safe, a national or regional policy decision should be made on feeding policy . Although breastfeeding transmission is reduced by ART, it is not abolished ,,[298-300],,. There is laboratory evidence that the breast milk of HIV-positive women on ART contains cells that may shed virus . As avoidance of breastfeeding can completely abolish the risk of postnatal transmission, this remains the recommended course of action.
There may be social or financial pressures on women to breastfeed, and support of formula feeding is important. The NSHPC report on perinatal HIV transmission in the UK  noted adverse social factors as a frequent factor in HIV transmission. A recent House of Lords report recommends the provision of free infant formula milk to HIV-positive mothers who have no recourse to public funds .
8.4.2 In very rare instances where a mother who is on effective HAART with a repeatedly undetectable VL chooses to breastfeed, this should not constitute grounds for automatic referral to child protection teams. Maternal HAART should be carefully monitored and continued until 1 week after all breastfeeding has ceased. Breastfeeding, except during the weaning period, should be exclusive and all breastfeeding, including the weaning period, should have been completed by the end of 6 months. Grading: 1B
Breastfeeding while not on HAART, or with detectable viraemia on HAART does constitute a potential child protection concern.
Because the risk of HIV transmission by breastfeeding is entirely avoidable, maternal breastfeeding against medical advice has previously been considered a child protection concern warranting referral to social services and, where necessary, legal intervention. The efficacy of ART in reducing HIV transmission by breastfeeding in the UK has not been measured. However, while the African data do not warrant a change in the recommendation not to breastfeed in these UK guidelines, they do make it likely that the risk of transmission is low enough that breastfeeding by a woman with HIV and fully suppressed virus on ART should no longer automatically constitute grounds for a child safeguarding referral. It is considered safer for women to be engaging with medical services while breastfeeding than for them to be breastfeeding without disclosing this. Data from Africa, in women not on HAART, show that mixed feeding carries a higher risk of HIV transmission than exclusive breastfeeding . It is recommended that breastfeeding be stopped as soon as is acceptable to the mother, but in any case by 6 months. A short period of mixed feeding may be necessary while ending breastfeeding.
8.4.3 Prolonged infant prophylaxis during the breastfeeding period, as opposed to maternal HAART, is not recommended. Grading: 1D
Studies in Africa have included both ART given to the mother and ART given as prophylaxis to the infant during breastfeeding. While serious adverse events were not reported in the infants given nevirapine for up to 6 months , there are currently insufficient safety data to advocate this approach given the particular safety concerns regarding the use of nevirapine in adults uninfected by HIV. The use of nevirapine for longer than the 2–4 weeks currently recommended for PEP is not advised .
8.4.4 Intensive support and monitoring of the mother and infant are recommended during any breastfeeding period, including monthly measurement of maternal HIV plasma VL, and monthly testing of the infant for HIV by PCR for HIV DNA or RNA (VL). Grading: 1D
Where a woman chooses to breastfeed against the medical advice in Recommendation 8.4.2, she and the baby should be monitored regularly for maternal adherence to ART; VL monitoring of the mother and diagnostic testing of the baby should be performed regularly (monthly). If the mother's adherence is suboptimal or she has detectable viraemia or an intercurrent illness that affects her ability to take or absorb ART, or she develops mastitis, she should be advised again to stop breastfeeding.
8.4.5 All infants born to mothers infected with HIV should have an antibody test at age 18 months. Grading: 1C
The potential for breastfeeding emphasizes the possibility of late transmission of HIV after the standard 3-month PCR test. Babies known to be breastfed should be tested monthly by PCR as above, but not all breastfeeding will be disclosed, and all babies born to HIV-positive women should have a negative HIV antibody test documented at age 18 months (see Section 8.5: Infant testing below).
8.5 Infant testing
8.5.1 HIV DNA PCR (or HIV RNA testing) should be performed on the following occasions (Grading: 1C):
- During the first 48 h and before hospital discharge.
- 2 weeks post infant prophylaxis (6 weeks of age).
- 2 months post infant prophylaxis (12 weeks of age).
- On other occasions if additional risk (e.g. breastfeeding).
HIV antibody testing for seroreversion should be checked at age 18 months.
8.6 Laboratory diagnosis of HIV infection in non-breastfed infants
The gold standard test for HIV infection in infancy was HIV DNA PCR on peripheral blood lymphocytes, although a number of studies, including the large French perinatal cohort have now demonstrated equal or increased early sensitivity with amplification of viral RNA with no false positives . Infants infected intrapartum may have low peripheral blood HIV levels, so HIV DNA/RNA may not be amplified from all infected infants at birth. Indeed a positive HIV DNA PCR result within 72 h of birth is taken as presumptive evidence of intrauterine transmission. Within the first few weeks of life, sensitivity of the viral diagnostic tests increases dramatically and by 3 months of age, 100% of non-breastfed HIV-positive infants are likely to be detected . In view of the genomic diversity of HIV where infant diagnosis will rely on HIV DNA amplification, a maternal sample should always be obtained for HIV DNA amplification with, or prior to, the first infant sample to confirm that the primers used detect the maternal virus. If the maternal virus cannot be detected then a different primer set and/or test should be used.
Infant HIV diagnostic testing should be undertaken at birth, 6 weeks and 12 weeks of age. Evidence from the French perinatal cohort demonstrated that neonatal ART, especially if more than one drug, can delay the detection of both HIV DNA and RNA in the infant . For this reason, the second and third HIV molecular tests are performed at 2 weeks and 2 months after stopping PEP (i.e. usually at 6 weeks and 12 weeks of age). If all tests are negative and the baby is not being/has not been breastfed, then parents can be informed that the child is not HIV infected. For infants at high risk of infection an additional early HIV test maybe undertaken at 2–3 weeks of age. For infants breastfeeding from mothers on HAART (see above), HIV viral diagnostic tests should be undertaken at least monthly on mother and infant while breastfeeding, and then twice on the infant, ideally between 2 and 8 weeks after weaning.
Loss of maternal HIV antibodies should be confirmed at 18–24 months of age. Ideally, an HIV antibody test should be used to confirm loss of maternal antibodies rather than a combined HIV antibody–antigen test. The latest tests are highly sensitive and may give a positive HIV result until up to 2 years of age . Testing for loss of maternal HIV antibody remains important as rarely, late postnatal infection may occur, even when all early HIV viral genome diagnostic tests were negative (French Perinatal cohort: five of 4539 cases) . This may be due to covert breastfeeding, premastication of infant food or unknown intrafamilial exposure.
If any of the infant HIV tests are found to be positive, an immediate repeat on a new sample should be requested to confirm infection. When an infant is found to be HIV positive, PCP prophylaxis should be started immediately, if the baby is not already on it, and an urgent referral to the local specialist HIV clinic should be made to initiate infant HAART. Maternal and infant HIV resistance testing should be undertaken to help delineate reasons for treatment failure and guide treatment. HIV services for children in the UK are organized in managed networks, details of the Children's HIV Network (CHIN) and contacts for local paediatricians can be found on the CHIVA website (http://www.chiva.org.uk) .
8.7 Child protection
Rarely, pregnant mothers refuse treatment for their own HIV as well as interventions to reduce the risk of transmission to their unborn infant. Whether for social, religious or other reasons, mothers who have been reluctant to accept interventions may be able to, where each aspect of the intervention package is dealt with separately (maternal ART, delivery, infant ART, infant feeding). This step-by-step approach has helped women to gradually make difficult personal changes to their birth plans. The input of the MDT is crucial to support these women, as they are often the most isolated and unsupported.
Where, despite all efforts, the MDT is unable to influence a mother's views antenatally, a pre-birth planning meeting with social services should be held. The mother should be informed that it is the paediatrician's role to advocate on behalf of the child's well-being and therefore to prevent, where possible, HIV infection. If the mother continues to refuse any intervention package, then legal permission should be sought at birth to treat the infant for 4 weeks with combination PEP and prevent breastfeeding. Preparation of the legal case may be lengthy and time consuming; useful documentation can be obtained from colleagues who have already undertaken this.