Alice Hoffsten: Necrotizing Enterocolitis in Preterm Infants: Impact on Infant Mortality and a Search for Predictive Biomarkers

Abstract

Necrotizing enterocolitis (NEC) is an intestinal disease with high mortality and morbidity. This thesis aims to identify biomarkers to possibly predict this fatal and increasingly common disease in preterm infants.

In Paper I, we studied cause of death (COD) in preterms treated at in the neonatal intensive care unit at Uppsala University Children’s Hospital, Uppsala, Sweden. From Period 1 (2002-2009; n=105) to Period 2 (2011-2018; n=160), NEC increased as COD. Paper I also found that the autopsy influenced COD determination in 34.9% of cases. In Paper II, we investigated if biomarkers at birth could identify preterms at risk of developing NEC. Blood samples were collected after birth from 40 preterms, gestational age (GA) <28 weeks. 11 infants later developed NEC (NEC-group), while 29 did not (Control-group). 189 biomarkers, reflecting inflammation, apoptosis and vascularisation, were quantified with Proximity Extension Assay (Olink®). No biomarker differed in expression when comparing the NEC-group to the Control-group, and only some biomarkers differed when compared to control sub-groups. These findings suggest that infants may not be predisposed at birth of developing NEC besides being preterm. In Paper III and IV, we analyzed intestinal biopsies from 43 infants operated due to NEC (NEC-group) and 27 infants (Control-group) operated due to conditions such as atresia, volvulus and aganglionosis. The biopsies were immunohistochemically stained for markers of Paneth Cells (DEFA6 and GUCA2A) in Paper III and Goblet cells (REG4) in Paper IV. Marker expression was quantified with a semi-automated digital image analysis (ImageJ®) and compared between the groups. Paper III revealed that the NEC-group had lower DEFA6 expression than the Control-group, while expression of GUCA2A was similar. In a logistic regression analysis, NEC-risk correlated to low DEFA6 expression. This suggests that NEC-infants have intact Paneth cells, as indicated by unaltered GUCA2A expression, but diminished DEFA6 activity. Paper IV showed that low REG4 expression correlated to higher NEC-risk. A multivariable logistic regression analysis with REG4 expression and GA, found that GA and not REG4 correlated to NEC-risk. Thus, low REG4 expression in NEC-patients may be maturity dependent. 

In conclusion, our findings suggest that preterms are not predisposed at birth to develop NEC besides being preterm. These studies also deepen our knowledge on Paneth cell and Goblet cell involvement in NEC. Low expression of DEFA6 and REG4 may indicate a risk of developing NEC, and could be further studied as potential blood biomarkers.