Monitoring Necrotizing Enterocolitis in Premature Infants

In this 5-year study, we extend our prior feasibility study to test the hypothesis that very-low-birth-weight neonates (VLBW) who develop necrotizing enterocolitis (NEC) can be reliably detected early in the process using broadband optical spectroscopy sensitive to changes in perfusion of the gut. Such perfusion changes result in regional tissue hypoxemia, which we have demonstrated to be detectable by combined broadband visible/near-infrared tissue oximetry, a real-time method that can assess the adequacy of deep tissue oxygenation. Initially this will be an optical-only device; then combining it with ultrasound will be studied in years 2-5, first on the benchtop, and then in the intensive care unit.

Necrotizing enterocolitis (NEC) is the most common life-threatening surgical emergency encountered in the neonatal intensive care unit [ ]. NEC is a multi-factorial (infectious, inflammatory, and ischemic) disease of the gastrointestinal (GI) tract of newborns and neonates. The end result is mucosal injury and/or transmural necrosis of the intestine. Currently, there is no test to diagnose NEC in the early stages of the disease, before the later and ominous clinical signs appear. The mortality of NEC ranges from 10% to 50%, approaching 100% for neonates with severe forms of the disease characterized by full-thickness necrosis of the intestine, followed by rupture and sepsis.

Ninety percent of NEC cases occur in infants born before 36 weeks' gestational age, occurring in up to 10% of all very-low-birth-weight (VLBW, <1500g) neonates [ ]. A diagnosis of NEC increases the NICU length of stay by 22-60 days, and increases the total hospital charges by $76,000-$186,000 per case [ ]. Finally, Neonates recovering from NEC often incur additional serious complications (malnutrition, liver dysfunction). NEC requiring surgery carries increased risk of cerebral palsy, cognitive or psychomotor impairment, or both.

Repeated attempts to use clinical signs to reliably identify neonates most likely to progress to severe NEC have been unsuccessful. Broadband oximetry appears to offer a solution. Developed by the PI and others, broadband oximetry is sensitive to ischemia. It differs from typical near-infrared spectroscopy (NIRS) methods that generally use only 2-4 wavelengths in that broadband oximetry measures hundreds of wavelengths. Broadband approaches, have shown significantly tighter normal ranges, lower noise, and better reproducibility in vivo. Further, typical NIRS fails to account for shifts in water, fat, blood volume, and stool, any of which can affect oxygenation measurements if not specifically accounted for, making NIRS unreliable for quantitative studies of the gut. In contrast, in our just-completed 1-year feasibility R43 trial, we demonstrated that NEC can be detected using broadband methods. The question remains: how will this new device perform clinically in a multicenter study? By incorporating broadband oximetry monitoring into the management of VLBW neonates, we may detect NEC at its earliest stages and prevent the cascade that leads to bowel necrosis. Success should lead quickly to clinical use, as this team has previously developed, received FDA approval for, and commercialized two biomedical devices.