Limited Versus Extended Trophic Feeding (LET-FEED) Trial

Up to 30% of all very preterm (VP; defined as those born before 32 weeks' gestation) newborns born within the United States develop late onset sepsis (LOS), a life-threatening infection that occurs after 72 hours from birth. LOS is the leading cause of morbidity in all very preterm newborns. VP newborns who develop LOS have significantly lower survival rates (adjusted risk ratio [aRR] 0.89, 95% confidence interval [CI] 0.87-0.90), 32% increased risk of being discharged on oxygen (aRR 1.32, 95% CI 1.26-1.38), 288% increased risk of needing a tracheostomy (aRR 2.88, 95% CI 2.47-3.37), and a 209% increased risk of requiring a gastrostomy tube (aRR 2.09, 95% CI 1.93-2.57).27 Thus, clinical practices that can reduce LOS are critically important to ensure VP newborns not only survive, but thrive, living lives free of disability or impairment.

One of the most protective factors in preventing LOS is an exclusive human milk-based diet. Common clinical practice for VP newborns includes starting human milk-based enteral feeds at approximately 20 mL/kg/day (e.g. "trophic feeds") and then progressively advancing the volume of enteral feeds daily until reaching a sufficient quantity to achieve adequate hydration and nutrition (e.g. "full feeds"). While enteral feeds are being advanced, VP newborns receive additional supplementary fluids and nutrition via intravenous (IV) infusions. Yet, any indwelling catheter to provide IV nutrition inherently increases the risk of LOS because the catheter breaches the newborn's protective skin barrier, thereby increasing risk of invasive infection from skin microbes.

One way to reduce the need for IV fluids is to initiate enteral human milk feeds earlier and advance feeds faster, thereby limiting the duration of indwelling IV catheters. An additional benefit of this approach is that early introduction of human milk prevents dysbiosis of the neonatal microbiome that is associated with LOS. By preventing the growth of harmful, pathogenic, sepsis-causing bacteria within the gut of VP newborns, these newborns have overall less risk of developing LOS. Yet, these benefits are with an exclusive human milk-based diet as meta-analyses and systematic reviews have demonstrated that feeding formula significantly increases the risk of necrotizing enterocolitis (NEC), a devastating inflammatory process within the gastrointestinal system that is associated with a 30-50% mortality rate.

A Gerber-funded (AA Salas PI, novice research award), pilot randomized trial that compared early versus delayed progressive feeding in 60 extremely preterm infants born at the University of Alabama Hospital showed that early progressive feeding reduced the need for central venous access (p=0.0001) and parenteral nutrition by 4 days (p=0.0005). Culture-proven sepsis (10% vs. 27%; p= 0.18) tended to be lower in the early progressive feeding group but did not reach statistical significance due to the limited power and sample size of the study.

Thus, the natural next question is whether early progression of enteral feeds in a multi-center trial improves clinical outcomes, namely prevention of invasive infection, without increased risk of adverse outcomes including mortality. The lack of multi-center clinical trials evaluating faster advancement of enteral feeds to full volume feeds has left clinicians with uncertainty about the risks and benefits of this approach and resulted in a lack of uniformity in clinical practice. This has led to significant differences in how VP newborns are fed across the US and globally, with some centers initiating and advancing feeds within the first few hours after birth and others continuing small volume "trophic" feeds without any advancement until day 4 after birth. Thus, clinicians caring for VP newborns are in dire need of evidence to determine the optimal feeding advancement strategy to improve their short- and long-term outcomes.

Study Population:

Investigators will enroll 350 very preterm infants born at 25 0/7 to 31 6/7 weeks' gestation admitted to the six participating neonatal intensive care units. All newborns will be eligible for this study without regards to gender, ethnicity, race, socioeconomic status, or language spoken. This study population has been selected based on the frequency of newborn admissions, sepsis, and overall mortality observed at these gestational ages. The participating centers include: (1) St. Joseph's Medical Center (Tacoma, WA), (2) Baylor College of Medicine (Houston, TX), (3) University of Alabama (Birmingham, AL), (4) University of South Florida (Tampa, FL), and (5) University of Oklahoma (Oklahoma City, OK).

Consent and Randomization:

To ensure timely enrollment, written informed consent will be obtained no later than 36 hours following birth, with a goal of enrollment within 24 hours of birth or prenatally. Identification of potential participants will prompt a visit from a study team member, who will meet with the parent(s) in their room or the baby's room to explain the study. During this meeting, the risks and benefits associated with participation will be thoroughly discussed, and ample time will be allocated for the parents to ask any questions they may have. The consent will include that the newborn will be fed parent's own milk as the prioritized enteral feed with donor milk as the alternative if POM is of insufficient quantity.

Importantly, participants will be assured that their infant will receive all necessary treatments irrespective of their involvement in the study. The process of randomization will then determine the assignment of each participant to their respective study group.

Participants will be randomly assigned to one of the study groups following computer generated random-block sequences. The University of Washington will develop the sequence and implement stratified randomization by gestational age using a computerized central telephone system. Twin and multiple infants will be randomized individually.

Study Intervention and Active Comparator:

After consenting, information about the pregnancy and participating infant will be collected, including contact information for the parent(s), race/ethnicity, medical history, family history, education level, employment status, marital status, and other information to assess baseline characteristics of study participants.

All consented and enrolled infants will be randomly assigned to receive either limited trophic feeds of 24 hours or extended trophic feeds of 72 hours after birth prior to the advancement of enteral feeds. Given that IV fluids and parenteral nutrition need to be ordered up to 12 hours prior to the implemented change, investigators have included a 12-hour period of advancement (24 to 36 hours versus 72 to 84 hours after birth) to account for these logistics and ensure the generalizability of the trial's findings. Feeding volumes will be advanced by 25 to 30 mL/kg/day until full feeds (140 mL/kg/day) are reached within two weeks of birth.

All participating centers will provide parent's own milk as the prioritized feed, with human donor milk provided inn cases where POM is not available in sufficient quantities. Preterm formula will not be used. The trial will recommend that fortification occurs once enteral feeding volumes of 40 to 60 mL/kg/day or higher are achieved and that IV catheters will be removed once the newborn is above 120 mL/kg/day of enteral feeds.

Deviations from the feeding protocol will be allowed if signs of feeding intolerance are present before establishment of full enteral feeding. Feeding intolerance will be defined as an interruption or cessation of enteral feeds for a period greater than 12 hours for abnormal abdominal examination within 14 days of birth. Information will also be collected on the percentage of enteral feeds that had a >50% reduction in planned volume over each 24-hour period to serve as an additional metric of feeding intolerance.

The primary outcome will be culture-proven late onset sepsis. Culture-proven LOS will be defined as positive blood, urine, and/or cerebrospinal fluid (CSF) cultures in the presence of compatible clinical signs of sepsis, occurring between postnatal day 3 and hospital discharge, necessitating treatment with antibiotics for 5 days or more. A subset of participants who agree to follow up will undergo 2 year developmental outcomes with the Bayley-4.