LAUSR

Prevention of late-onset neonatal sepsis (LOS) in infants born very preterm (<32 weeks’ gestation) is paramount because of its associated mortality and morbidity. Lactoferrin has antimicrobial properties and could be preventive. Lactoferrin concentrations in human breastmilk vary little with gestational age and fall over time after birth; they are lower in infant formulas. Importantly, very preterm infants have low enteral intakes in the first days after birth, and this can persist for weeks. Consequently, infants born very preterm might benefit from supplemental lactoferrin to prevent LOS. In a 2017 Cochrane review, enteral lactoferrin was found not only to reduce the rate of any LOS (risk ratio [RR] 0·59, 95% CI 0·40–0·87; p=0·008; from six trials including 886 participants) but also of necrotising enterocolitis (0·40, 0·18–0·86; p=0·02; from four trials including 750 participants). Five randomised controlled trials (RCTs) assessed bovine lactoferrin and one tested human recombinant lactoferrin. However, the quality of all the included studies was low. The authors of the review made no recommendation to introduce bovine lactoferrin into clinical practice. In The Lancet, the ELFIN Trial Investigators Group reports its large multicentre RCT, which was designed to address the limitations of the existing evidence in the use of bovine lactoferrin to prevent LOS. They recruited 2203 infants born at less than 32 weeks’ gestation and younger than 72 h from 37 sites around the UK in just under 3·5 years. Infants were randomly allocated to receive either bovine lactoferrin (150 mg/kg daily, once the infants were receiving at least 12 mL/kg per day enterally) or an equal volume of sucrose placebo until 34 weeks’ gestation. Bovine lactoferrin had little effect on the primary endpoint of LOS (316 [29%] of 1093 infants in the lactoferrin group and 334 [31%] of 1089 in the control group developed LOS; adjusted RR 0·95, 95% CI 0·86–1·04; p=0·233). Importantly, lactoferrin had little effect on other clinically important outcomes including mortality (1·05, 0·66–1·68; p=0·782) and the combined outcome of mortality with LOS, necrotising enterocolitis, or bronchopulmonary dysplasia (1·01, 0·94–1·08; p=0·743). The study design and governance appear exemplary. Infants were at equal risk of reaching the endpoint through randomisation, other clinical care was equal because of effective blinding, the primary clinically important outcome of LOS was known for all but 1% (n=23) of those enrolled, the infants were representative of very preterm infants in high-income countries, and treatment compliance was excellent. Moreover, they reported other clinically important outcomes including mortality, necrotising enterocolitis, bronchopulmonary dysplasia, and retinopathy of prematurity, and the combined outcome of death with any LOS, necrotising enterocolitis, bronchopulmonary Published Online January 8, 2019 http://dx.doi.org/10.1016/ S0140-6736(18)32390-0