Steven Abrams, MD

1
Infant formula safety and the safety and efficacy
of donor human milk

• Steven Abrams, MD
• Professor of Pediatrics
• USDA/ARS CNRC
• 1100 Bates St.
• Houston TX 77030
• Ph713-798-7124 FAX [phone]
• sabrams_at_bcm.edu

2
Overview and objectives

• Infant formula safety
• Objective Understand basis of regulation of
  infant formula content in US
• Objective Recognize rationale for concern
  regarding infectious risks of infant formulas and
  methods of limiting these risks
• Banked donor human milk
• Objective Recognize rationale for the use of
  banked donor human milk for at-risk infants
• Objective Understand methods of assuring safety
  of these products
• Objective Recognize areas of ongoing research

3
Case study Missing formula components

• In Oct. 2003 three infants were hospitalized in
  Israel within a week of each other with
  encephalopathy.
• In Nov. 2003, a 5.5-month-old infant was admitted
  to a medical center in Israel with nystagmus and
  vomiting.
• Infant was treated with thiamine and improved
  within hours.
• All had been receiving the same soy-based
  formula. Ultimately the use of this formula in
  Israel resulted in 2 deaths and 13 additional
  hospitalizations.
• The formula was found to have 1/10 the labeled
  amount of thiamine (29-37µg vitamin B-1 per 100 g
  formula vs. the labeled 385 µg per 100 g).

Siegel-Itzkovich , BMJ, 2003
4
Recent recall (Sept. 2006)
5
Also
6
Assurance of nutritional adequacy of infant
formulas in the US

• Requirements for nutrients in infant formulas are
  established by the FDA in the Federal Food, Drug,
  and Cosmetic Act.
• Nutrient specifications include minimum amounts
  for 29 nutrients and maximum amounts for 9 of
  those nutrients.

http//www.fda.gov/opacom/laws/fdcact/fdcact4.htm
7
Assurance of nutritional adequacy of infant
formulas in the US

• If formula does not contain nutrients at or above
  minimum level or within range, it is considered
  an adulterated product unless the formula is
  "exempt" from certain nutrient requirements.
• Manufacturer analyze representative samples for
  all nutrients at least every 3 mo.
• Manufacturers must notify FDA after processing or
  formulation changes as well as for new
  formulations.
• Regulations for infant formulas last revised in
  1985.
• Infant Formula Evaluating the Safety of New
  Ingredients (NAS, 2004). Process for evaluating
  new ingredients suggested. Decision tree analysis
  proposed for new ingredients.

http//www.fda.gov/opacom/laws/fdcact/fdcact4.htm
and http//www.cfsan.fda.gov/lrd/cfr106.html
8
Dont forget to keep the government and the
public informed!
The food company Nutricia has been warned by
food authorities in Australia and New Zealand
that they may have breached the Australia New
Zealand Food Standards Code by placing infant
formulas on the market before receiving necessary
approvals. While the Authority is not aware
of any evidence that the Nutricia products
containing FOS are unsafe, the products have not
been through the Food Standards Australia New
Zealand (FSANZ) approval process which includes
a safety assessment. FOS is permitted
within prescribed limits in infant formula sold
within the European Union. Products identical to
Nutricia Karicare Gold Plus Infant and Follow-On
formulas have been sold in Europe since 2000 with
no known health issues among babies or infants.
These products have been on sale in Australia
since January 2007 and NSW Health is not aware of
any reported problems. Nutricia have been
informed that they should have sought express
permission from Food Standards Australia New
Zealand (FSANZ) before adding any unapproved
nutritive substance to their infant formulas.
http//www.foodauthority.nsw.gov.au/advice-to-pare
nts-and-caregivers-nutricia.asp
9
Enterobacter sakazakii Case Study

• A 1.3 kg infant was delivered at 33 wks via
  c-section.
• DOL 11 - fever, tachycardia, and possible
  seizures
• CSF culture grew E. sakazakii - infant died.
• Additional suspected cases (tracheal
  colonization) and non-symptomatic infants in same
  NICU.
• Single batch of a specific powdered infant
  formula product (Portagen) was found to be
  significantly associated with E. sakazakii
  infection or colonization
• Opened and unopened cases from a single batch of
  Portagen grew E. sakazakii.

Morbidity and Mortality Weekly Report, CDC, 2002
10
Epidemiology

• E. sakazakii implicated in 1958 in 2 cases of
  neonatal meningitis in England.
• Worldwide, since 1958 there have been 70
  reported cases (24 deaths) of E. sakazakii
  infection believed acquired via powdered infant
  formula.
• Full-term infants lt28 days (meningitis),
  premature infants (sepsis) are at greatest risk.
• Reported mortality rates between 33-80.
• Muytjens (1990) examined 141 different infant
  formulas from 35 different countries and isolated
  E. sakazakii from 20 formula samples from 13
  countries.
• Current Codex Alimentarius Commission
  specifications for E. sakazakii permit 1-10
  CFU/gram of powdered infant formula.

Drudy , CID, 2006 Gurtler Int J Food Microbiol,
2005
11
E. sakazakii cases
Drudy et al., CID, 2006 Gurtler et al., Int J
Food Microbiol, 2005 Boween and Braden, Emerging
Infectious Diseases 2006121185.
12
General guidelines for the reconstitution of
powdered infant formula (PIF) in hospitals

• PIF should have clear indication for use.
• Personnel should use aseptic techniques to
  prepare PIF.
• The use of sterile liquid formula is encouraged
  for healthy newborns. Single-use bottles,
  nipples, feeders recommended. All mixing
  utensils, bottles, and nipples should be
  disinfected
• Formula prepared in advance should be done daily,
  and should be kept at lt 4C and discarded if not
  used in 24 hours.
• Hang time for continuous feeds should be not
  over 4 hours. Intermittent feedings should be
  prepared as a single feeding.

Drudy et al., CID, 2006 Gurtler et al, Int J
Food Microbiol, 2005, ADA 2002.
13
General guidelines for the reconstitution of
powdered infant formula (PIF) in homes

• Families of high-risk, formula-fed infants should
  be alerted that PIF is not a sterile product and
  encouraged to use liquid formula when possible.
• Inactivation of E. sakazakii in infant formula
  will occur at temperatures 60C or higher.
• European society (ESPGAN) recommends against this
  due to potential nutrient loss. Use of water at
  this temperature has not been included in US or
  TCH guidelines.
• Detailed specific guidelines for
  high-risk/preterm infants do not exist. Limit PIF
  through 44 weeks PMA?

Agostoni et al, J Pediatr Gastroenterol Nutr,
2004 Drudy et al., CID, 2006 Gurtler et al, Int
J Food Microbiol, 2005
14
Boiling - effects on nutrients
No meaningful changes for Vit A, D,E,K,B1,B2,B6
Per label tap water boiled, cooled to 40C then
added and shaken. Boiling water boiled water
put into bottle and power added immediately.

• http//www.fao.org/ag/agn/jemra/enterobacter_en.st
  m

15
Guides for parents
http//www2.texaschildrenshospital.org/internetart
icles/uploadedfiles/142.pdf
16
Safety and efficacy of using banked human milk in
the hospital setting
17
Outline

• Background and potential benefits to human milk
  banking using own mothers milk (OMM) or donor
  milk (DM) in a high-risk population.
• Guidelines for donor human milk banking
• Safety Procedures
• Potential limitations
• Clinical Evidence
• Conclusions

18
Benefits of human milk for at-risk infants

• There are many well-recognized benefits of human
  milk for term, and (when fortified), premature
  infants
• Fortified human milk provides nutrients with high
  bioavailability and supports optimal growth
• Human milk reduces incidence and severity of
  infections, especially necrotizing enterocolitis.
• Human milk-fed infants have a lower incidence of
  atopic disease and possibly other chronic
  illnesses.
• Infants fed human milk tend to have improved
  development, especially, preterm infants
• Despite advances in infant formula, human milk is
  the optimal nutrient source for nearly all
  infants.

AAP. Pediatrics, 2005115496-506.
19
Bioactive factors in human milk Food for the
brain protection for the GI Tract

• Secretory IgA
• Lactoferrin
• Enzymes (e.g., PAF-acetylhydrolase)
• Cytokines (e.g., IL-10)
• Growth factors (e.g., EGF)
• Nucleotides
• Oligosaccharides
• Antioxidants
• Glutamine
• Polyunsaturated fatty acids
• Micronutrients

Hamosh M. Pediatric Clin No Amer 20014869
20
Necrotizing enterocolitis in preterm infants (UK)
In-Hospital Diet All Cases Confirmed Cases
Formulas Only (n236) 24 (10.2) 17
(7.2) Formulas plus Mothers Milk (n437) 16
(3.7) 11 (2.5) Human Milk (n 253) 11 (4.3) 3
(1.2)
OR Formulas only vs Human Milk only all cases
2.5 (1.25.2), plt0.02 confirmed 6.5 (1.922),
plt0.001 OR Formulas only vs Formulas as
supplements to Mothers Milk all cases 3.0
(1.55.7), plt0.005 confirmed 3.0 (1.46.5),
plt0.005 Multi-center Not randomized Human Milk
Mothers own milk and/or pasteurized donor milk
Lucas Cole, Lancet 19903361519-23
21
Survival Curves for NEC or death by amount of
human milk (ml/kg/d)
1.00
100 ml
0.95
Survival
50 ml
Estimate
0.90
20 ml
10 ml
0.85
0 ml
For NEC or Death after 14 days, adjusted for
birth weight, race, PDA treatment, ventilation,
and site Meinzen-Derr, et al NICHD Neonatal
Network. Presented SPR 2006
0.80
0
10
20
30
40
50
60
70
80
90
100
110
120
Postnatal age (d)
22
Effect of human milk on URI symptoms in preterm
infants during first year
Human Milk
Formula
BW 1.2 kg BW 1.2 kg GA 29 wk GA 29 wk DC 60
days DC 51 days
60
Cumulative URI Days (mean)

40

20


0
From discharge to
1 mo
3 mo
7 mo
12 mo
Human Milk (any)
100 71 29 10
Bier, J Perinatol 200222354-359
23
Human milk and subsequent intellectual
performance in premature infants at 8 y
Significant factors affecting IQ
Social Class - 3.5/class Mothers Education
2.0/group Female Gender 4.2 Mechanical
Ventilation - 2.6/week Receipt of Human Milk
8.3 IQ points
Lucas, Lancet 1992339261
24

Outcomes by quintiles of breast milk intake in
extremely premature infants
Breast Milk Percentile
None ? 20th 20- 40- 60- gt 80th
40th 60th 80th

Breast milk, mL/kg/d 0 1 7 24 64 111 DC on
breast milk () 0 1 2 9 40 85 Mean MDI
score _at_ 18 mo 76 74 77 78 80 87a Mean PDI score
_at_ 18 mo 81 81 83 84 84 89a Mean BRS ile _at_ 18
mo 46 45 52 50 52 59a Re-hospitalization lt1y
() 30 25 32 26 23 13a
Vohr, (Neonatal Newtwork) PEDIATRICS 2006 118
e115 avalue significantly different from
None Bayley Scales of Infant Development II
N775 Breastmilk (75) 260 No Breastmilk (25)
A. MDI, Mental Development Index Birth
weight 800 g Gestational age 27 wk B. PDI,
Psychomotor Development Index C. BRS, Behavior
Rating Scale
Co-variates Mat age, education, marital
status,race, GA, gender, infection, IVH, PVL,
BPD, NEC, weight
25
Magnitude of breast milk effect

• For every 10 mL/kg/day ? breast milk intake
• MDI ? 0.5 pts
• PDI ? 0.6 pts
• Behavior Rating Scale percentile score ? 0.8 pts
• Re-hospitalization rate ? 6
• For the highest quintile, an intake of 100
  mL/kg/day
• ? 5 pts on MDI

Vohr, PEDIATRICS 2006 118 e115. Hack, NEJM
2002 346149
26
Effect of predominant mothers milk feeding on
late onset sepsis (LOS) and necrotizing
enterocolitis (NEC) in premature infants
Preterm Formula
50
50
Mothers Milk
45
45
Fortified Mothers Milk gt50 mL/kg/day
40
40

• LOS/NEC p lt 0.01
• Schanler, PEDIATRICS 19991031150

35
35
Infants
30
30
25
25
20
20
15
15
10
10
5
5
0
0
Late onset Necrotizing sepsis
enterocolitis
27
Human milk fortification

• A response to the need to provide additional
  nutrients, especially minerals, for premature
  infants
• Protein
• Calcium
• Phosphorus
• Zinc
• Used for most infants lt 1.8 -2.0 kg
• Fortifiers available
• Similac Human Milk Fortifier (Ross)
• Enfamil Human Milk Fortifier (Mead Johnson)

28
Incremental effect of fortification
100
600
90
Increase over Unfortified HM
Increase over Unfortified HM
80
70
400
60
50
40
200
30
20
10
0
0
Ca P Zn Vitamins
Energy Protein Na Fat CHO
29
Fortified vs unfortified human milk

• gt 600 infants randomized
• Growth
• Weight gain (g/kg/d) 3.6 2.74.6
• Length (cm/wk) 0.12 0.07 0.18
• Head circumference (cm/wk) 0.12 0.07 0.16
• Bone mineral content (mg/cm) 8.3 3.8 12.8
• Nitrogen balance (mg/kg/d) 66 35 97
• BUN (mg/dL) 16 8 24
• Relative Risk
• Feeding intolerance 2.9 0.6 13 NS
• Necrotizing enterocolitis 1.3 0.7 2.5 NS
• Death 1.5 0.7 3.3 NS

Kuschel CA Harding JE 2005 The Cochrane
Library Some comparisons with partial
supplements
30
Donor human milk

• Has the benefit of providing species-specific
  passive immunity.
• Protection against necrotizing enterocolitis in
  premature infants (Mcguire W, 2003).
• Offers minimal risk of feeding intolerance.
• Has been used for infants with metabolic
  disorders, chronic renal failure, failure to
  thrive, and short gut syndrome.
• Sources are non-commercial and commercial milk
  banks.

31
Donor milk banks Non-commercial
www.hmbana.org
1 Canadian and 10 US milk banks belong to the
Human Milk Banking Association of North America
(HMBANA).
32
Guidelines for donor milk banking in the US

• As members, all voluntarily follow the Guidelines
  for the Establishment and Operation of a Donor
  Human Milk Bank, generated by HMBANA.
• Guidelines include protocols for soliciting
  donors and collecting, processing and
  distributing milk.
• All donors screened for antibodies to HIV-1,
  HIV-2, HTLV-1, HTLV-2, HBs Ag, Hep C, and
  syphilis.
• Donors must also provide a full health and risk
  history and PPD if appropriate.

33
Tully, J Hum Lact, 200218393-6.
Donor milk supply and costs

• Donor milk is dispensed either on hospital
  purchase order or physician order for a baby not
  hospitalized.
• Donor milk banks are non-profit. All milk is
  donated.
• Cost associated with donor screening and milk
  processing is offset by a processing fee
  (3.50/oz in the US for HMBANA milk) which is
  usually paid by the ordering institution or the
  recipient. Some insurance policies cover this.

34
Donor qualifications
35
Commercial donor milk bank

• Milk donations obtained from healthy donors (not
  paid) after medical history. Donor is screened
  for markers to HBV, HCV, HIV-1 and 2, HLTV-1 and
  2, and syphilis. The milk is matched with the
  donor using DNA fingerprinting.
• Each lot of milk is screened for the presence of
  bacterial contamination, drugs, and non-human
  proteins.
• A sample analyzed for calories and
  macronutrients. Milk is either centrifuged to
  separate the fat or ultra-filtrated to
  concentrate proteins. The skim and cream are
  blended to bring the milk macronutrients to
  target levels.

36
Commercial milk banking procedures

• Milk is pasteurized using a High Temp Short Time
  (HTST) process (standard temp 57C-72C).
• After pasteurization, milk is stored at -30C.
• Pathogen log-reduction of HIV is gt 7.3
• Similar results for other bacteria and viruses
• Methods similar to those used in preparation of
  human blood products.

37
Donor milk banking Efficacy?
38
Efficacy of donor milk Effects of pasteurization
Source commercial donor milk company
39
NEC and donor milk Meta-analyses
RR 0.2595 CI 0.06 to 0.98. McGuire Arch Dis
Child Fetal Neonatal Ed 2003

• Another recent meta-analyses, using data from gt
  20 yrs ago found virtually identical results (RR
  0.21 95 CI 0.06 to 0.76). Boyd et al. Arch Dis
  Child. Fetal Neonatal Ed. Published online April
  5, 2006.

40
NEC and donor milk if include recent study
Morales and Schanler. Sem Perinatol 20073183-8.
41
Trial comparing mothers milk, donor milk or
formula Recent study

• Randomized, blinded trial comparing
  infection-related events among extremely preterm
  infants fed mothers milk, donor human milk, or
  preterm formula
• Infants lt30 weeks gestation were randomly
  assigned to receive pasteurized donor milk or
  preterm formula if their own supply of mothers
  milk became insufficient during the study

Schanler, et al., Pediatrics, 2005116400-6.
42
Results

• No differences between group DM and group PF in
  incidence of infection related events, duration
  of hospital stay, or number of deaths. Compared
  with groups DM and PF, MM had fewer infection
  related events, and shorter hospital stay.

Schanler, et al., Pediatrics, 2005116400-6.
43
Conclusions

• Limitations
• DM and preterm formula groups received up to half
  their OMM which may have obscured differences.
  21 of infants switched over from donor milk to
  preterm formula.
• Study not powered to detect differences in NEC.
  Incidence of NEC was much lower than pre-study
  expectation.
• Conclusions
• Larger studies are needed to determine benefits
  of donor human milk. Ideal comparison donor
  human milk vs preterm formula in infants who
  never received OMM.
• However, providing donor milk to infants whod
  previously received human milk did not improve
  outcome relative to infant formula when OMM was
  no longer available.
• Potential cost savings to banked human milk, if
  decreases incidence of NEC (not tested here),
  would be considerable.

44
Donor milk and formula nutrient contents
Component Donor Formula HM (per dL)
Milk Comm-fortifier
Energy (kcal) 60 87 87 Fat (g) 2.2 4.0 3.8 Prote
in (g) 1.1 2.2 2.2 Na (mEq) 0.6 1.5 1.3 Ca
(mg) 26 78 73 P (mg) 11 45 39
Donor Milk Banked, pooled donor milk, not
pasteurized. Tyson, J. Pediatr 1983 10395
www.prolacta.com
45
Donor milk conclusions

• Most studies gt 20 years old and larger babies
• Most donor milk was pasteurized removing most
  immunologically-active components. This is less
  true now.
• NEC, sepsis not primary outcomes/not powered for
  these. Newer products can be adapted to higher
  nutrient intakes
• It is currently difficult to evaluate benefits of
  banked HM. Key outcome variables of feeding
  tolerance and decreasing risk of NEC require
  large multi-center studies.
• Evidence of safety is excellent
• Cost-benefit analysis is difficult at this time
  until more data are available Currently, donor
  human milk is inadequately addressed in public
  health policy.

46
Objectives revisited

• Infant formula safety
• Objective Understand basis of regulation of
  infant formula content in US.
• Objective Recognize rationale for concern
  regarding infectious risks of infant formulas and
  methods of limiting these risks
• Banked human milk
• Objective Recognize rationale for the use of
  banked human milk for at-risk infants
• Objective Understand methods of assuring safety
  of these products
• Objective Recognize areas of ongoing research so
  as to be able to follow this rapidly-developing
  topic

47
Thanks