I. BACKGROUND. Necrotizing enterocolitis (NEC) is the most common and most serious gastrointestinal (GI) emergency of the neonate. Its pathogenesis is complex and multifactorial, and the etiology remains unclear. In spite of the advances in neonatology over the last few decades, the mortality and morbidity secondary to NEC remains high. Current clinical practice is directed mainly toward prompt, early diagnosis and institution of proper intensive care management.

A. Definition. NEC is an acute inflammatory injury of the distal small and often proximal large intestine. Surgical pathology reveals segmental coagulative necrosis of the mucosa with focal hemorrhage as evidence for ischemia. Other features include intramural gas (pneumatosis) and sloughing of mucosa, submucosa, and muscularis mucosa, which is in contrast to the preserved mucosal integrity in spontaneous intestinal perforation. Universally accepted risk factors include prematurity, bacterial dysbiosis, and formula feeding.

B. Epidemiology. Despite decades of research, NEC remains the most common serious surgical disorder among infants in a neonatal intensive care unit (NICU) and is a significant cause of neonatal morbidity and mortality.

1. The incidence of NEC varies from center to center and from year to year within centers. There are endemic and epidemic occurrences. An estimated 0.3 to 2.4 cases occur in every 1,000 live births. In most centers, NEC occurs between 1% and 5% of all NICU admissions and 5% to 10% of very low birth weight (VLBW) infants. Mortality ranges from 20% to 40% but can approach 100% in case of NEC totalis. Overall, NEC is responsible for 12% of deaths in extreme premature infants <27 weeks of gestational age.

2. At least 30% of NEC cases result in surgical resection of affected tissue. However, the timing for surgical intervention and the type of surgery remain controversial. Severe NEC that requires surgical intervention increases the average length of stay by 43 days and is associated with increased morbidity (e.g., short bowel syndrome) and mortality.

3. Prematurity is the single greatest risk factor. Decreasing gestational age is associated with an increased risk of NEC. The postnatal age at onset is inversely related to birth weight and gestational age, with a mean age at onset of 12 days. The mean postmenstrual age of infants with NEC is between 30 and 32 weeks.

4. Approximately 10% of infants with NEC are term. Risk factors for this population include congenital heart disease with presumed decreased intestinal perfusion (e.g., hypoplastic left heart syndrome, coarctation of the aorta), polycythemia, intrauterine cocaine exposure, and intestinal anomalies such as gastroschisis. The colon appears to be the most commonly affected site.

5. NEC pathogenesis is multifactorial. Additional risk factors for the disease include histologic plus clinical chorioamnionitis, IUGR, maternal smoking, polycythemia, and other—less well supported—maternal or neonatal conditions. Antenatal steroids improve the maturity of the GI tract and have been shown to reduce the incidence of NEC.

6. Recently, prolonged empiric antimicrobial use has been associated with increased NEC occurrence and matches other studies that documented decreased microbial diversity and overgrowth of potential pathogenic bacteria (dysbiosis) prior to clinical presentation of NEC.

7. Although bacteria are clearly involved in the pathogenesis of the disease, no single infectious organisms have been isolated except in relative rare outbreak situations. NEC should be differentiated from infectious (viral) or allergic (milk intolerance) colitis.

8. Transfusion-associated NEC (TANEC) has been described in numerous reports and although blood transfusions were independent risk factors in several retrospective studies, a causal relationship has not yet been confirmed in larger prospective trials.

9. A rare, more benign form of NEC has been described: pneumatosis coli. Neonates without the typical risk factors for NEC present with grossly
bloody stools, minimal or absent abdominal and systemic signs, and isolated colonic pneumatosis without small bowel involvement.

10. Almost all infants with NEC have received enteral feedings prior to disease onset. Formula feeding increases the risk of NEC (relative risk is 2.8). However, up to 6% of infants <1,250 g birth weight still develop the disease despite receiving exclusively breast milk.

1. The pathogenesis of NEC remains a conundrum. NEC is a multifactorial disease resulting from complex interactions between immaturity, mucosal injury, and bacterial imbalances. Because these factors affect most preterm infants, infants who develop NEC must also exhibit an especially harmful inflammatory response to intestinal antigens.

2. Genetic polymorphisms have been described in patients at higher risk for severe NEC such as in genes encoding toll-like receptor (TLR) 4 or interleukin 18 (IL-18) signaling. Polymorphism in the secretor gene fucosyltransferase (FUT) 2 encoding low secretor status have been associated with earlier and more severe disease.

3. Intestinal immaturity plays an important role in the pathogenesis of NEC: increased permeability of the intestinal epithelium, decreased motility, a thinner mucus layer, low or absent levels of secretory IgA, and lack of regulatory adaptation of the intestinal mucosal immune system.

4. Experiments in germ-free animals and TLR4 knockout mice strongly suggest that bacterial antigen is critical for the initiation of intestinal inflammation and NEC development. Previous studies have been hampered by the inability of conventional microbial culture techniques to thoroughly characterize the human GI microbiota because 80% of bacteria colonizing humans are not detected by culture methods. Sequencing the 16S small subunit bacterial ribosomal RNA (16S rRNA) gene or the entire bacterial genome discovered reduction in microbial community diversity with a shift toward increased abundance of potentially pathogenic subgroups. Although not representative for the vast majority of sporadic NEC cases, the literature contains numerous reports of NEC “outbreaks” with detection of various specific bacteria or viruses.

5. An excessive and inappropriate intestinal inflammatory response appears to be the key inciting event that leads to NEC. Although specific antigenic triggers may vary, failure to downregulate the innate immune receptor TLR4 on intestinal epithelial cells and lower ratios of FOXP3⁺ T regulatory cells in the mucosa are examples that can explain why the poorly adapted premature intestine is prone to inflammatory injury. Prenatal factors (e.g., corticosteroids vs. chorioamnionitis) may influence the “inflammatory set-up” of the preterm gut at birth.

6. Evidence supports a critical role for inflammatory mediators. Platelet activating factor (PAF), bacterial endotoxin, lipopolysaccharide (LPS), tumor necrosis factor (TNF), proinflammatory interleukins, and nitric oxide are some of the inflammatory mediators that have been studied in the pathophysiology of NEC. Both animal studies and samples from human infants demonstrate the association of elevated levels of PAF
in infants with NEC compared with those without. In animal models, exogenous administration of PAF mimics NEC-like injury and PAF antagonists limit such injury. Various other inflammatory mediators such as cyclooxygenase (COX-2), reactive oxygen species, tumor necrosis factor alpha (TNF-α), and IL-18 have been implicated in NEC pathogenesis mostly in animal models.

These data also point to the multifactorial etiology of the disease and underline the fact that not one but several strategies are necessary for the prevention of NEC.

7. Several retrospective studies have suggested a temporal association of packed red blood cell (PRBC) transfusions with the onset of NEC, but this link remains to be confirmed by prospective matched case-control studies. Hypothesized mechanisms include T cell activation, immune-mediated hemolysis, and splanchnic vasoconstriction associated with the transfusion. However, it remains unclear if blood transfusions play a causal role for NEC or are just indicators of impending or ongoing disease.

8. A large number of other factors such as low Apgar scores, timing and volumes of feeding, umbilical catheterization, hypoxic-ischemic insults, presence of a patent ductus arteriosus (PDA), or treatment with indomethacin or vasopressors have not been uniformly confirmed as independent pathophysiologic contributors.

II. DIAGNOSIS. Early diagnosis of NEC may be an important factor in determining outcome. This is accomplished by a high index of suspicion and careful clinical observation for nonspecific signs in infants at risk.

A. Clinical characteristics. There is a broad spectrum of disease manifestations. The clinical features of NEC can be divided into systemic and abdominal signs. Most infants have a combination of both, although abdominal signs usually predominate.

1. Systemic signs. Respiratory distress, apnea and/or bradycardia, lethargy, temperature instability, irritability, poor feeding, hypotension (shock), decreased peripheral perfusion, acidosis, oliguria, bleeding diathesis

2. Abdominal (enteric) signs. Abdominal distension or tenderness, gastric aspirates (feeding residuals), vomiting (of bile, blood, or both), ileus (decreased or absent bowel sounds), hematochezia (grossly bloody stools), abdominal wall erythema or induration, persistent localized abdominal mass, or ascites

3. The course of the disease varies among infants. Most frequently, it will appear (i) as a fulminant, rapidly progressive presentation of signs consistent with intestinal necrosis and sepsis or (ii) as a slow, paroxysmal presentation of abdominal distension, ileus, and possible infection. The latter course will vary with the rapidity of therapeutic intervention and require consistent monitoring and anticipatory evaluation (see section III).

B. Laboratory features. The diagnosis is suspected from clinical presentation but must be confirmed by diagnostic radiographs, surgery, or autopsy.
No laboratory tests are specific for NEC; nevertheless, some tests are valuable in confirming diagnostic impressions.

1. Imaging studies. The abdominal radiograph will often reveal an abnormal gas pattern consistent with ileus. Both anteroposterior (AP) and cross-table lateral or left lateral decubitus views should be included. These films may reveal bowel wall edema, a fixed-position loop on serial studies, the appearance of a mass, pneumatosis intestinalis (the radiologic hallmark used to confirm the diagnosis), gasless abdomen indicating ascites, portal or hepatic venous air, pneumobilia, or pneumoperitoneum with the appearance of gas under the diaphragm. Of note, extremely low birth weight (ELBW) infants often present with abdominal distension and ileus because intramural gas or pneumoperitoneum become more commonly presenting features after 30 weeks’ postmenstrual age. Spontaneous intestinal perforation (SIP) may present with pneumoperitoneum without other clinical signs.

Abdominal ultrasound can be a more sensitive method to detect intramural air and portal venous gas in experienced hands. Doppler studies can confirm bowel necrosis by absent blood flow. These techniques are particularly helpful to confirm radiographic appearance of pneumatosis intestinalis in well-appearing infants with feeding intolerance.

2. Blood and serum studies. Thrombocytopenia, persistent metabolic acidosis, and severe refractory hyponatremia constitute the most common triad of signs and help to confirm the diagnosis. Serial measurements of C-reactive protein (CRP) may also be helpful in the diagnosis and assessment of response to therapy of severe NEC. Blood cultures are positive in ˜40% of cases.

3. Analysis of stool for blood has been used to detect infants with NEC based on changes in intestinal integrity. Although grossly bloody stools may be an indication of NEC, routine testing of stool for occult blood has no value for NEC diagnosis. Approximately 60% of infants will have Hemoccult-positive stools at any given time during hospitalization without any evidence for NEC.