Necrotizing enterocolitis - classification and two initial steps towards prevention.

The premature infant suffers from immaturity of all organ systems, one of them being the gastrointestinal tract. When the infant is born, the immature gastrointestinal tract is exposed to milk and simultaneously colonized by high densities of bacteria. The combination of milk, microbiota and an immature gut, leaves the infant vulnerable to developing the dreaded intestinal emergency necrotizing enterocolitis (NEC). NEC is often very aggressive and no cure exists, which means that prevention is an utmost important topic to researchers, physicians, parents - and infants.   Despite immense research during the last decades, no specific test to determine if an infant suffers from NEC exists. Most neonatal units use Bell's staging criteria, which are based on clinical and radiographic findings, as a diagnostic tool; however the diagnosis given according to Bell's stages has not been validated. In study I, we aimed to determine the validity of the NEC diagnosis given at discharge. An expert panel consisting of a neonatologist, a paediatric surgeon and a paediatric radiologist served as the golden standard. We found that the diagnosis given at discharge had a poor validity which significantly affected the reported incidence of NEC in the neonatal department at Rigshospitalet, Denmark. The validity of the NEC diagnosis was worse than the validity of most other paediatric diagnoses that had been investigated.   In studies II and III, we aimed to explore possible means of NEC prevention. The role of nutrition in NEC development is well established with mother's milk as the best option to avoid NEC in the preterm infant. Maternal milk is, however, most often not available in sufficient amounts during the first days of life, and preterm infant formula or human donor milk is used in its absence. Studies in preterm piglets showed that bovine colostrum equally to human donor milk protected against NEC compared to infant formula. Furthermore, bovine colostrum was superior to human donor milk in stimulating gut immunity and digestive functions.   Hence, in study II we aimed to design a pilot study of bovine colostrum used as a supplement to maternal milk in the first days of life and to determine if the study was feasible. In the paper, we present the protocol and the results of the first two phases of the Precolos study in which 12 infants were included and received pasteurized, spray-dried and reconstituted bovine colostrum during the first days of life as the first infants in the world. We found that the infants tolerated bovine colostrum without clinical adverse effects, but we also observed a transient hypertyrosinemia on day seven of life in five infants. The results were evaluated by a safety management board which encouraged us to continue the pilot study with the last phase, which was a randomized controlled trial of 20+20 infants comparing supplementation with bovine colostrum to supplementation with standard nutrition. The randomized trial has just finished recruitment.   At last, we wanted to shed light on a possible microbiological angle of NEC prevention. Dysbiosis and bacterial translocation are believed to play a crucial role in the development of NEC as intestinal pneumatosis, which occurs when bacteria produce gas inside the intestinal wall, is a pathognomonic radiographic finding. In a quality improvement study from the US published in 2014, NEC incidence was significantly reduced after the implementation of several quality improvement interventions. Standardized weekly exchange of nasogastric feeding tubes was suggested as one of the potential NEC-reducing interventions.   In the neonatal unit at Rigshospitalet, Denmark, preterm infants are fed 8-12 times daily through a resident nasogastric feeding tube which is exposed to body temperature, contains milk residuals from the last meal and is handled by both parents and personnel. Since bacterial pollution of milk given through the nasogastric feeding tube might be NEC-inducing, we aimed in study III to determine the bacterial load given to the infants when feeding them through a tube. We collected 92 used nasogastric feeding tubes and flushed them with one ml saline each to imitate a meal given through them. Eighty-nine percent of the tubes contaminated the meals with more than 1000 colony-forming units of bacteria and fifty-five percent contaminated the meals with the possible pathogens Enterobacteriaceae or Staphylococcus aureus. The concentration of bacteria in the saline flushed through the tubes was as high as 109 colony-forming units per ml; however, neither the risk of contamination nor the concentration of bacteria in the flush was associated with the duration of use. Implementation of standardized weekly exchange of feeding tubes would therefore not prevent the contamination of meals.   In conclusion, the studies included in this thesis serve as a base for future studies investigating the prevention of NEC. We found a poor validity of the NEC diagnosis given at discharge. This should be kept in mind when conducting epidemiological studies of NEC and especially when conducting interventional trials with NEC as an outcome. If the findings of the randomized part of the Precolos study indicate a positive effect of bovine colostrum and do not give rise to concerns regarding feasibility, safety or tolerability, a large-scale randomized controlled study with NEC as the primary outcome will be planned. Based on the high concentrations of bacteria found in the nasogastric feeding tubes, a randomized controlled trial investigating whether the frequency of feeding tube exchange affects the early colonization has been commenced in the neonatal department at Rigshospitalet. Hopefully, the results of these studies will bring us closer to preventing NEC in the future.