The aim of the project is to study the effects of fortification (using a Human Milk Donor Fortifier) of an exclusive preterm human milk diet on outcome of extremely preterm neonates, born at less or equal to 27 weeks.

II. BACKGROUND Necrotizing enterocolitis (NEC) is a serious gastrointestinal disease affecting >10% of extremely preterm infants with a mortality of up to 50%. The pathophysiology is poorly understood, but intestinal immaturity, microbial colonization, hypoxic ischemia, and diet have all been implicated. Use of human milk and standardized feedings are the only strategies that consistently have been shown to decrease the incidence of disease. Our NICU has been using a standardized feeding protocol with an emphasis on human milk for over 20 years. In 1995, Crouse became a NY state licensed human milk bank, which accepts donors that have delivered preterm. Preterm breast milk has distinct nutritional advantages over milk obtained from donors who delivered term infants. In particular, there is a higher protein and mineral concentration in preterm human milk compared to mature term milk which is what is typically available in all commercial milk banks. All extremely preterm infants in our NICU receive preterm human milk - maternal or donor. Combined with following our standard feeding protocol, this has resulted in >90% less NEC than in comparable NICUs. The investigators reported an incidence of NEC of 1% over a recent 6 year period, compared to an incidence of 12% among the three other upstate NY regional perinatal programs (Albany, Rochester and Buffalo).

While the prevention of feeding related morbidities such as NEC is paramount, growth remains a concern in these extremely preterm infants. Growth velocity has been shown to have a significant effect on neurodevelopmental outcomes in at least one study. Human milk fortifiers are marketed and used to various degrees in an effort to increase weight gain in VLBW infants fed maternal breast milk (MBM). Our current feeding protocol calls for "selective" human milk fortification with a human milk derived fortifier (HMDF; Prolact+4 H2MF, Prolacta Bioscience, Morovia, CA). Our criterion for fortification is sustained weight gain <15 g/kg/day. Most other centers use fortification more routinely. The impact of fortification on feeding related morbidities (e.g. NEC) as well as growth and longer term outcomes are unknown. For example, one recent randomized trial included infants at high risk for NEC (birth weight <1250 g). All of the infants in that study were fed their own mother's milk but differed by the type of supplementation used. Some infants received HMDF (Prolacta) and donor human milk if maternal milk became unavailable (exclusive human milk diet), whereas others received a bovine milk-based fortifier with maternal milk and preterm formula if no maternal milk was available (bovine group). In the exclusive human milk group, some infants were assigned randomly to receive fortification early (when milk intake was only 40 mL/kg/day) and some later (when milk intake was 100 mL/kg/day). While the incidence of NEC was reduced from 26% for infants in the bovine group to 7% for those in the exclusive human milk group , the incidence was still seven-fold higher than that reported by us. Also of note, the group who received the earlier fortification had a higher incidence of NEC than the group that received later fortification (8% vs 6%). It is unclear what role early fortification plays in the pathogenesis of NEC, but theoretical concerns include effects on host-defense properties and milk osmolality.

In summary:

1. Exclusive use of human milk is associated with a lower rate of NEC.
2. Preterm human milk (maternal or preterm donor) has distinct advantages over commercial donor human milk that comes from mothers of mature term infants.
3. Fostering more rapid growth seems advantageous, but the role of fortification in the development of NEC and other morbidities is unclear as are the long term benefits.

Therefore, the purpose of our study is to compare an exclusive preterm human milk diet using routine early fortification with a diet that uses fortification selectively for poor growth.

III. SPECIFIC OBJECTIVES

1. To compare the effectiveness of early routine human milk nutritional fortification vs. selective fortification on in-hospital growth in extremely preterm infants.
2. To compare the incidence of feeding intolerance and NEC in extremely preterm infants receiving early routine fortification vs. those receiving selective fortification for poor weight gain.
3. To compare the effect of early routine human milk nutritional fortification vs. selective fortification on growth after discharge from the NICU up to 2 years of age.
4. To compare the effect of early routine human milk nutritional fortification vs. selective fortification on neurodevelopmental outcomes at 2 years of age.

IV. STUDY DESIGN

This is a randomized controlled trial in extremely preterm infants, which will be conducted at the Crouse Hospital NICU in Syracuse, NY. Parents of eligible infants will be approached by study investigators for informed consent prior to the initiation of enteral feeds. Enrolled infants will then be randomized by milk bank personnel into two groups:

1. Early routine fortification (RF): HMDF added to preterm human milk (maternal or donor) when feeds of 150 mL/kg/day are reached.
2. Selective fortification (SF): HMDF added to preterm human milk only for weight gain <15 g/kg/d after full feeds of 180 mL/kg/day are achieved.

Full feeds (120Kcal/kg/day) is achieved with unfortified milk at a volume of 180 ml/kg/day and with fortified human milk at a volume of 150ml/kg/day.

Enteral feedings will be prepared by the bedside nurse. The primary outcome will be growth velocity (g/kg/day) after regaining birth weight. Our prior study of ELBW infants in the Crouse NICU, showed a mean growth velocity of 13 ± 2.5 g/kg/day. Two-hundred study subjects would adequately power the study to show a 1 g/kg/day difference between study groups (α=0.05, β=0.2). Based on previous admission rates to our NICU, this enrollment could be achieved in 18 months.

Secondary outcomes will include NEC, time to full feeds (surrogate for feeding intolerance), other growth parameters (length and head circumference), and neurodevelopmental outcomes at 24 months of age, adjusted for prematurity.

V. SUBJECT SELECTION Inclusion Criteria: This study will include all extremely preterm infants with gestational age at less or equal to 27 weeks who are admitted to the NICU prior to day of life seven and the first enteral feeding.

Exclusion Criteria: Infants with severe congenital anomalies such as chromosomal trisomy, congenital heart disease, abdominal wall defect will be excluded, and severe Intra Ventricular Hemorrhage prior to entering the study.

VI. STATISTICAL METHODS, DATA ANALYSIS, AND INTERPRETATION Parametric data will be analyzed using the student t-test and presented with means and standard deviations. Non-parametric data will be analyzed using the Mann-Whitney U-test and presented with medians and interquartile ranges (IQR). Categorical data will be analyzed using the Chi-squared or Fischer's exact test. Weight growth velocities will be calculated using an exponential model validated by Patel et al. from regained birth weight to discharge. Absolute measures of growth (weight, length, head circumference) will be converted to Z-scores using a research calculator (http://www.ucalgary.ca/fenton) based on parameters from the Fenton Preterm Growth Chart.

Z-scores for growth parameters at six-months and 24 months will be calculated based on data from the World Health Organization growth charts (http://www.peditools.org/growthwho/index.php).

VII. STUDY PROCEDURES Feeding Protocol Both parenteral nutrition and enteral feedings for this study will be based on protocols that have long been used in our NICU and published. Parenteral nutrition will be provided to both the RF and the SF groups in the same way. Total parenteral nutrition (TPN) will be started on the day of birth with at least 1 g/kg/day of protein (Premasol) and will be advanced as tolerated to a standard solution. The standard TPN solution delivers 100 kcal/kg/day in 150 mL/kg/day, and contains 18.8 g/kg glucose, 3 g/kg protein, and 2.5-3.0 g/kg lipid (Intralipid).

Initiation of trophic feeds with MBM or preterm donor milk (PDM) will be started seven days after birth and continued at 12 mL/kg/day for one week in both groups. Infants are fed by gastric gavage every two hours until reaching a weight of 1250 g and every three hours thereafter.

After trophic feedings are completed, milk intake is advanced by 15 mL/kg/day as tolerated to a volume of 150 mL/kg/d. Infants in the RF group will be started on Prolacta HMDF to achieve milk with 24 kcal/oz and maintained at an intake of 150 mL/kg/day (120 kcal/kg/day). Infants in the SF group will have human milk increased to 180 mL/kg/day (120 kcal/kg/day) and maintained at this volume unless fortification is required. If weight gain is <15 g/kg/day after one week of intake at 180 mL/kg/day, fortification will be started in the SF group and feeds will be decreased to 150 mL/kg/day. If weight gain remains < 15g/kg/day after one week of fortified milk at 150 ml/kg/day in either the RF or SF groups, the intake will be increased to 165ml/kg/day ( 132 kcal/kg/day). Intralipid and parenteral nutrition will be discontinued when enteral feedings reach 80 and 100 mL/kg/day, respectively.

Per our NICU's current standard of care, feedings will be decreased or stopped for emesis, bile stained gastric residuals, or abdominal distension. Feedings will not be advanced if the previous day's residuals exceeded 10% of daily volume. If feeds are stopped, they will be restarted without fortification; if fortification was being used prior to a stoppage, it will be resumed once the infant is tolerating feeds of 150 mL/kg/day.

In the case of a blood transfusion, feedings will be modified as per our standard of care. Feeds will be cut back to 100-120 mL/kg/day and fortification will be stopped in both study groups. Following the transfusion, feeds will be advanced by up to 20 mL/kg/day. Fortification can be resumed 24 hours after the blood transfusion is completed.

Infants in the study will have daily weight checks, along with weekly recumbent length measurements and head circumference measurements. These growth parameters will be followed until 35 weeks adjusted age, which is the earliest infants can be discharged home from our NICU. In case infants are transferred to an outlying NICU prior to discharge, growth parameters will continue to be followed at the accepting hospital. After 35 weeks adjusted age, infants will be fed maternal milk if available or a preterm infant formula.

Data Collection In addition to the growth data described above, baseline patient demographics, clinical characteristics, nutrition information, and NICU outcomes will be collected on all patients. The demographics and characteristics will include birth weight, gestational age, sex, race, inborn/outborn status, receipt of maternal betamethasone, mode of delivery (vaginal or cesarean section), and Apgar score at 5 minutes. Nutritional information will include day of first enteral feeding, type of first enteral feeding (MBM, PDM, combination), duration of trophic feedings (< 20 mL/kg/day), duration of TPN (days), duration of need for intravenous fluids (days), days to full enteral feedings, enteral feeding at full feeds (MBM, PDM, combination), time of human milk fortification (days of life), and enteral feeding at 35 weeks adjusted for prematurity (MBM, PDM, combination). NICU outcomes will include grade III/IV intraventricular hemorrhage (IVH), duration of mechanical ventilation, surfactant therapy, use of postnatal steroids, bronchopulmonary dysplasia (BPD), nosocomial infection, spontaneous intestinal perforation (SIP), necrotizing enterocolitis (NEC), length of stay, number of blood product transfusions, and discharge disposition.

Follow-Up Infants will be seen in our NICU Follow-up clinic at six months, 15 months, and 24 months adjusted age, which is our current standard of care. At these visits, a health history including hospitalizations and diet will be collected. Growth parameters-weight, length, and head circumference-will be measured and a neurodevelopmental assessment using the Bayley Scales of Infant and Toddler Development will be administered. Developmental assessments are conducted by staff blinded to prior feeding history.

Data & Safety An independent safety monitoring committee composed of an attending neonatologist (Dr. Ellen Bifano) and milk bank personnel will meet after the first 20 patients are enrolled and periodically to monitor safety during the study. Specifically, feeding related morbidities like NEC will be followed.

All study data will be compiled into an electronic database, which will be kept in a secure password-protected computer in the principal investigator's locked office on the ninth floor of Crouse Hospital. The study subjects' personal health information including name, date of birth, and medical record numbers will be used to identify subjects, and their study information will be part of their medical record. A copy of the data collection form and data analysis form will be submitted along with this protocol and application.

Inclusion Criteria:

• All extremely preterm infants with gestational age less or equal to 27 weeks who are admitted to the NICU prior to day of life seven and the first enteral feeding

Exclusion Criteria:

• Infants with severe congenital anomalies such as chromosomal trisomy, congenital heart disease, and abdominal wall defect