Infant Feeding

1
Infant Feeding HIV/AIDS New data and
recommendations for action

• Ellen G. Piwoz, ScD
• Africas Health in 2010
• Academy for Educational Development
• Presentation to the ACCESS Project
• September 12, 2006

2
Overview

• Background
• New data on infant feeding HIV/AIDS
• ARVs breastfeeding (WHO consultation May)
• Infant feeding options
• Early breastfeeding cessation
• Implications for programs

3
Timing of Mother-to-Child Transmission No
intervention
Early Antenatal (lt36 wks)
Late Postpartum (6-24 months)
Early Postpartum (0-6 months)
Labor and Delivery
Breastfeeding
Pregnancy
Late Antenatal (36 wks to labor)
5-10
10-20
5-20
4
Risk Factors For Postnatal Transmission of HIV

• Infant
• Age (first month -?)
• Breastfeeding duration
• Non-exclusive BF
• Lesions in mouth, intestine
• Low birth weight
• Genetic factors host/virus

• Mother
• Immune/health status
• Plasma viral load
• Breast milk viral load
• Breast inflammation (mastitis, abscess, bleeding
  nipples)
• New HIV infection

WHO, 2004
5
Causes of death in children lt 5 yrs

Deaths associated with under-nutrition 50-60
Sources EIP/WHO, Caulfield LE, Black RE. Year
2000
6
New data on infant feeding and HIV/AIDS

• ARV use and breastfeeding

7
Combination short-course ARV prophylaxis regimens
reduce IU, IP early PN transmission
significantly to 4-7 in breastfeeding
populations at 6 wks
As IU IP transmission decrease, the
contribution of breastfeeding as a cause of
pediatric HIV infection will increase
8
Data from a study in Kenya report that
single-dose NVP suppresses breast milk HIV-1 RNA
in first 3-21 days post-delivery Chung et al,
AIDS, 2005

• 76 Kenyan mothers randomized to receive sd-NVP
  vs. ZDV
• ZDVThai/CDC regimen 2x daily from 34 wks
  every 3 hr during labor (no PN ZDV)
• BM samples taken 2-4 times/week from birth-weeks
  (795 samples from 60 mothers)
• All mothers reported exclusive breastfeeding
• Transmission at 6 wks
• 6.8 for NVP vs. 30.3 for ZDV (p0.02)

9
Studies are underway to examine the efficacy of
daily infant ARV prophylaxis to prevent BF
transmission

• To date, the impact appears to be comparable to
  short-course ZDV/NVP /-3TC _at_ 6 wks. Transmission
  at 6 mo influenced by shorter BF duration.

10
Research is also looking at the impact of
maternal HAART on postnatal HIV transmission

• Two distinct questions
• What is the safety efficacy of maternal HAART
  for reducing BF-associated transmission among
  women who meet criteria for starting treatment?
  (20-35 of women)
• Provide treatment to those who need it compare
  outcomes to those who are not as sick (and
  receiving short course prophylaxis)
• Is maternal HAART more efficacious than
  short-course regimens for preventing
  BF-associated HIV-transmission in women who do
  not meet criteria for treatment?

11
Women who meet criteria for HAART have a 2 to 4
fold greater risk of MTCT even when receiving
short-course ARV prophylaxis (data from Cote
dIvoire Trials Data, F. Dabis 6/05 BF 50)
WHO Criteria for ART WHO Stage 4 or Stage 3 and
CD4lt350 or Stage 1-2 and CD4lt200
12
Mothers who are treatment-eligible are also more
likely to transmit HIV during breastfeeding
Iliff et al, AIDS, 2005
gt80 of PNT by mothers with B/L CD4 gt 350
occurred gt 6 months
13
After introduction of HAART for eligible
mothers, there was no difference in HIV
transmission to 18 mo among infants who were
formula fed (FF) vs. breastfed (BF) from birth -
Botswana(Thior et al, JAMA, 2006)
Note Median BF duration 5.8 months Findings
were similar for mortality, and infections
death
14
Recommendations from WHO Consensus Meeting
regarding ARV use and breastfeeding

• Current UN recommendations regarding infant
  feeding apply to women receiving HAART
• If avoidance of BF is AFASS, then RF birth
  otherwise EBF
• Studies are looking at potential toxicity from
  infant exposure to HAART via breast milk but the
  known risks of RF from birth presently outweigh
  theoretical concerns
• 2. No evidence to support provision of either
  maternal HAART or prolonged infant prophylaxis
  solely for the purpose of preventing
  breastfeeding-associated HIV transmission
• Studies are underway.

New ARV guidelines recommendations
disseminated at the IAS meeting in Toronto
15
New data on infant feeding and HIV/AIDS

• Infant feeding options
• Breast vs. formula feeding

16
Botswana MASHI Study feeding interventionThior
et al, JAMA, 2006

• 1200 women randomized to

Formula feed with 1 month ZDV prophylaxis
(government standard)
Exclusively breastfeed with extended infant ZDV
prophylaxis through 6 months
vs.

• RESULTS
• HIV infection at 7 mo 9.1 in BF gt 5.6 in
  FF (p0.04)
• Infant mortality at 7 mo 4.9 in BF lt 9.3 in
  FF (p0.004)
• HIV death at 18 mo 14.2 in FF, 15.6 in BF
  (p0.41)

17
MASHI Study -continuedThior et al, JAMA, 2006

• Although HIV transmission lower in FF, early
  mortality is greater. Thus, no difference in
  HIV-free survival was observed for FF vs. BF (6
  mo) infants at 18 mo.
• FF infants had higher incidence of adverse
  events, hospitalization, and non-HIV related
  infectious diseases
• After HAART introduced, no difference in PNT,
  mortality, HIV-free survival for FF BF
• Risks associated with formula feeding outside of
  research settings may be even greater

18
Not breastfeeding was associated with increased
risk of hospitalization and death during a
diarrhea disease outbreak in Botswana -1 T Creek,
CDC, PEPFAR 2006 (tgc0_at_cdc.gov)

• HIV prevalence in pregnant women 33.4 (2005)
• PMTCT program provides free infant formula for 12
  mo to HIV-exposed infants
• 63 HIV mothers chose formula feeding from birth
  (2005)
• 35 of all infants lt 6 mo not BF (including
  infants born to HIV-negative mothers of unknown
  HIV status)
• Nov 2005-Feb 2006 heavy rains public water
  supply contaminated (rural)
• 35,046 cases of diarrhea 532 deaths in Q1
  (Jan-Mar) 2006

Photo from CDC
19
Risk factors for diarrhea among children
presenting at emergency room -2 T Creek, CDC,
PEPFAR 2006, (tgc0_at_cdc.gov)
Models adjusted for SES, age, and mothers HIV
status Not BF associated with 8.5 fold
increased risk of mortality. 51 infants who
died did not receive adequate supply of infant
formula predisposed to malnutrition
20
Cumulative risk of postnatal HIV infection at 18
mo (95 CI) in BF FF infants exposed to
different ARV prophylaxis regimens DITRAME Plus
Leroy et al, IAS Conference, 2006
107/926 children HIV-infected of whom 27 were PNT
cases
(16-30)
(10-27)
(6-14)
(4-11)
(2-10)
PNT Cases
15
10
1
1
0
21
18 month probability of survival was similar for
formula fed and breastfed infants (3-4 mo)
infants in Abidjan

• Incidence of severe adverse events mortality
  were similar for FF SDBF after adjusting for
  infant HIV status
• Both interventions were superior to prolonged BF
  historical cohort (plt0.0001)
• Given appropriate nutritional counseling and
  care, access to clean water, and supply of breast
  milk substitutes, these alternatives to prolonged
  breastfeeding can be safe in urban African
  settings.

Becquet et al, IAS, 2006 infected infants 27
SDBF, 23 FF, 51 Prolonged BF
22
Longer breastfeeding duration was associated with
reduced risk of mortality among infants born to
HIV mothers in rural Uganda Homsey et al,
preliminary data, PEPFAR conference, Durban 2006

• Preliminary analysis - 52 babies born to mothers
  on HAART 28 babies born to mothers not yet
  eligible for HAART
• Median BF duration 3.2 mo HAART 5.1 mo
  non-HAART
• HIV transmission n1 in HAART group n5 in
  non-HAART infants
• Infant mortality 25.7/100 person-years in HAART
  group 7.7/100 PY in non-HAART group
• Each additional month of BF associated with a 38
  reduction in infant mortality after adjusting for
  maternal HAART infant infection status
  (p0.004)
• Adjusted HR 0.72 (95 CI 0.6- 0.9)

23
Promoting exclusive breastfeeding in the general
population may reduce HIV transmission by HIV
mothers who do not know their infection status -
via reductions in mixed BFPiwoz et al, AJPH, in
press
N365 p0.04 in test for trend. Each additional
intervention contact was associated with a 38
reduction in PNT after adjusting for maternal CD4
24
Early introduction of solid foods was associated
with increased risk of HIV transmission -
Vertical Transmission Study (VTS), South Africa
(Coovadia, et al, IAS, 2006)
25
Even though relatively higher SE mothers were
more likely to choose RF from birth, the risk of
mortality associated with RF was greater than the
risk associated with BF VTS
(Coovadia, et al, IAS, 2006)
2.5-fold increased risk of death associated
with RF in this predominantly rural peri-urban
environment (82)
26
New data on early breastfeeding cessation
replacement feeding after 6 months

• Compliance
• Risks
• Options

27
Compliance with early breastfeeding cessation in
Abidjan (Becquet et al, JAIDS, 2005)

• Mothers counseled to stop BF _at_ 3-4 mo 38
  compliance
• Problems associated with early cessation
• Breast pain 33
• Engorgement 18
• Infant crying 17
• Family disapproval 2
• Living with in-laws having CD4gt 500 were
  associated with continued BF (plt0.04)
• Poor feeding practices were associated with
  higher risk of growth faltering (stunting)

28
Zambia Exclusive Breastfeeding Study (ZEBS)-
Early rapid breastfeeding cessation ( _at_4 mo)
increases the risk of breast health problems
reduces duration of amenorrhea Thea et al, AIDS,
2006
Note The prevalence of breast health problems
among mothers who stopped BF is comparable to
other studies (HIV, non-HIV)
29
BM viral load increased after early rapid
breastfeeding cessation placing infants at
increased risk if BF is resumed Thea et al, AIDS,
2006
BM virus detectable _at_ gt 50 copies/mL data for
those with detectable virus only No increase
in BM VL seen in women who continued BF
30
Diarrhea malnutrition are other possible
consequences of early breastfeeding cessation
placing infants at increased risk of subsequent
hospitalization death--a multi-study analysis
of the impact of early breastfeeding cessation
underway
31
In food insecure populations, cost poor diet
quality make it impossible to meet the
nutritional needs of non-breastfed children using
locally consumed foods Mozambique Linear
Programming example (6-8 mo) Alons et al, PEPFAR
meeting, 2006

• Replacing BM with local foods increases the cost
  of feeding by 3-fold
• Food scarcity is seasonal (drought) and
  widespread (42)
• Existing diet is low in several nutrients (iron,
  zinc, vitamin C)
• Need for fortified foods increased feeding
  frequency dietary diversity to prevent
  malnutrition associated with EBC

Diet without breast milk 5000 MT(0.20)
Sorghum maize flour, GLV, banana, coconut oil,
cows milk
32
Is early breastfeeding cessation by HIV mothers
AFASS in rural Zimbabwe? Tavengwa, et al, IAS,
2006

• Formative study to understand barriers
  facilitating factors
• Carried out in context of PMTCT program that
  provided infant PCR test results at 5 months
• Many mothers motivated to stop BF to avoid
  infecting their infants but few planned for how
  they would feed their infants after BF stopped
• Economic factors, lack of disclosure, and stigma
  associated with not BF were major barriers

• Maybe one of the most contributing factors is
  that they would not have disclosed, so to
  abruptly wean would mean that a lot of social
  problems are going to arise, maybe the in-laws
  are going to ask why mother has weaned the baby
  and for the mother to say I have weaned because I
  am positive I want to save my baby from HIV
  infection it will be difficult. Health Worker

33
Cumulative risk of PNT to 24 mo was lt 3 in
mothers with plasma VL lt 10,000 copies m/L at
delivery(n1256) Taha et al, Toronto IAS
meeting, TUPE0337
25.9
12.3
4.2
2.6
The risk of PNT was 1.2 from 1.5-6 months in
infants exposed to NVP. Weaning by 6 mo could
prevent gt 85 of late PNT. However to minimize
post-weaning morbidity mortality, continuation
of BF may be considered if maternal plasma VL is
low
34
The story to now It appears that the best
approach to reducing PNT is to screen mothers for
treatment eligibilityHAART should be provided
to those who are eligibleShort-course ARV
prophylaxis should be provided to other
mothersInfant feeding counseling should still
be a matter of personal choice
35
The story to now (cond) Formula feeding
from birth does not appear to convey added
benefits in settings where there are risks of
malnutrition, morbidity mortality from unsafe
preparationGreater support for exclusive
breastfeeding is warranted (particularly to
reduce mixed feeding)Many unanswered
questions regarding the timing, safety, impact
of early breastfeeding cessation Low-cost,
nutritious, bacteria-resistant replacement foods
are needed to optimize growth, health, and
survival true for all, not just HIV-exposed
36
Priority Actions for HIV Infant Feeding
WHO/UNICEF 2004

• Develop/revise comprehensive IYCF policy that
  includes HIV infant feeding
• Implement/enforce International Code on Marketing
  of BMS
• Intensify efforts to promote, protect, support
  appropriate IYCF practices in the general
  population
• Provide adequate support to HIV-positive women to
  make carry out their feeding decisions
• Support operational research, ME and learning at
  all levels

37
Some strategies that work

• Adapting global IF guidelines to local context
• Focus on feeding children 6 mo-24 mo
  Mozambique, Zimbabwe, Zambia
• Supporting adequate IF training for health
  providers in MCH PMTCT with follow-up/refresher
  courses, quality monitoring
• Incorporating IF counseling/support into
  post-natal care
• Counseling should extend beyond pre-natal visits
• Strengthening follow-up support for EBC RF
• Challenging, so need to test different strategies
• Engagement of communities mentoring of
  community support groups
• Baby-friendly communities
• Strengthen ME of infant feeding within PMTCT
  reporting
• Suggested indicators in new WHO PMTCT strategy
• Increase access to treatment for HAART-eligible
  mothers
• Monitor HAART eligibility throughout the PN period