OVERVIEW: What every practitioner needs to know

Are you sure your patient has necrotizing enterocolitis (NEC)? What are the typical findings for this disease?

Necrotizing enterocolitis (NEC) is an acute inflammatory disease of the gastrointestinal tract of preterm infants that was initially described in the 1950s. The pathology of NEC can range from limited mucosal injury to full thickness necrosis of the small and large intestine. The distribution of injury may be focal, segmental, multifocal, or diffuse. The three most common signs of NEC are as follows:

• Abdominal distention (See Figure 1)

• Bloody stools (grossly bloody, or heme-positive for occult blood); and

Continue Reading

• Excessive gastric residuals (fed infants) or bilious gastric aspirates (unfed infants).

Figure 1.

Clinical and gross intestinal pathological features of NEC

Other signs referable to gastrointestinal pathology include vomiting, diminished to absent bowel sounds with ileus, abdominal tenderness, cellulitis of the abdominal wall, and less commonly, a palpable abdominal mass.

The effects of the acute inflammatory process in NEC typically extend beyond the gastrointestinal tract. Systemic signs such as lethargy, new onset or increased episodes of apnea, new onset of respiratory distress or worsening of respiratory distress, poor color, diminished peripheral perfusion, bleeding diasthesis, hypotension, bradycardia, and shock syndrome are almost always observed at some stage in the disease process. These systemic signs may occur even if cultures of blood and other body fluids are sterile.

Pathology similar to NEC in preterm infants has occurred in term infants with other conditions such as cyanotic congenital heart disease, Hirschsprung’s disease, imperforate anus, intestinal atresia, and severe perinatal asphyxia. However, the etiology and pathogenesis of intestinal injury in these situations likely differs from that of classic NEC in preterm infants.

Isolated intestinal perforation (IIP) is a distinct entity that results from a punctate perforation of the gastrointestinal tract. IIP can present with marked abdominal distention due to a pneumoperitoneum and systemic signs of sepsis due to intraperitoneal spillage of fecal organisms. This entity has been associated with medications such as indomethacin and steroids.

What other disease or condition shares some of these symptoms?

The clinical presentation of NEC ranges from an insidious progression over several days that may begin with nonspecific systemic signs, to a fulminant onset in which gastrointestinal signs and shock develop over hours.

The differential diagnosis includes bacterial or viral sepsis and/or meningitis; causes of anatomic or functional intestinal obstruction (e.g., malrotation, intestinal atresia, intussusception, aganglionosis); conditions that can result in critically low intestinal blood flow (e.g., intestinal volvulus, hypoplastic left heart syndrome, critical coarctation of the aorta); omphalitis; isolated intestinal perforation (IIP); and in term infants, milk protein intolerance.

What caused this disease to develop at this time?

NEC is a disease that occurs almost exclusively in preterm infants with postmenstrual ages less than 35 (and usually less than 32) weeks. The incidence of NEC in infants with birth weights between 500 and 1500 grams ranges from 3% to 10% in various large published series and increases with decreasing weight. In addition to prematurity, well-established risk factors for NEC include overly aggressive enteral formula feedings and colonization of the bowel with pathogenic bacteria. These risk factors are supported both by epidemiologic studies and investigations using animal models.

Preterm infants demonstrate immature absorptive, digestive, and immunologic functions and diminished peristaltic activity. In this setting, even cautious advancement of enteral feedings can cause milk stasis and intestinal dilatation. These in turn can initiate an excessive inflammatory response by the immature intestinal epithelium, especially in the presence of colonization by an abnormal profile of microbiota or a specific enteric pathogen. Local gastrointestinal inflammation can trigger a wider systemic inflammatory response that may not be distinguishable clinically from sepsis.

Other risk factors for NEC in preterm infants, such as perinatal asphyxia, catheterization of umbilical vessels, respiratory distress syndrome, apnea and bradycardia, administration of blood transfusion, anemia and hypoxemia have not been substantiated in prospectively designed and well-controlled investigations.

Small for gestational age infants, especially those with microcolon, may be at increased risk for NEC.

Recent evidence suggests that treatment with antibiotics is a risk factor for NEC, probably by altering the distribution of intestinal flora.

What laboratory studies should you request to help confirm the diagnosis? How should you interpret the results?

There is not one laboratory study that is diagnostic of NEC. Rather, a diagnosis of NEC can be supported by a constellation of laboratory abnormalities that include the following:

Complete blood count. The absolute number of white cells may be elevated but is more usually markedly depressed, especially as the disease progresses. Neutropenia and an increased ratio of immature to mature neutrophils (a left shift) are common. Hematocrit typically decreases due to a bleeding in the injured gastrointestinal tract. Thrombocytopenia (platelet count < 100,000/μL) occurs due to increased platelet consumption and depressed bone marrow production.

Blood culture. 30-50% of infants with NEC develop culture-proven bloodstream infection. It is important to remember that sepsis without gastrointestinal inflammation can mimic the presentation of NEC (due to the development of a secondary ileus).

Serum sodium. Hyponatremia will result when there is significant “third spacing” of extracellular fluid and loss of sodium within the gastrointestinal tract.

Arterial blood gases and serum bicarbonate. Metabolic acidosis characterized by a low serum bicarbonate is an ominous sign that may indicate tissue hypoperfusion.

Serum glucose. Hyperglycemia (serum glouse > 150 mg/dL) often develops as a secondary stress response.

C-reactive protein (CRP) and procalcitonin. Concentrations of these inflammatory markers are usually elevated.

Coagulation abnormalities. Infants with advanced NEC develop laboratory evidence of a coagulopathy, including elevations in PT and PTT and reductions in serum fibrinogen, suggestive of a disseminated intravascular coagulation process.

Culture, chemistries, cell counts, and gram stains of fluid obtained by paracentesis of a fluid pocket under ultrasound guidance will identify peritonitis that is suggestive of occult intestinal perforation or transmural necrosis with translocation of bacterial or fungal pathogens.

Would imaging studies be helpful? If so, which ones?

Initial and serial abdominal radiographs should be obtained, including the traditional KUB study as well as left lateral decubitus (left side down) and/or cross-table lateral examinations as appropriate.

Findings that are pathognomonic of NEC are pneumatosis intestinalis (See Figure 2); microscopic and gross anatomical appearance (See Figure 3) and portal venous gas (See Figure 4). Pneumatosis refers to the presence of gas within the intestinal wall. “Bubbly” pneumatosis occurs when intestinal gas (including hydrogen gas that is produced by bacterial fermentation of enteral feedings) dissects into the intestinal mucosa. “Linear” pneumatosis is evident when gas has dissected beneath the subserosa. Portal venous gas can be seen as linear lucencies over the liver that are due to migration of intraluminal gas into the mesenteric capillaries and subsequent drainage of mesenteric venous blood into the portal system.

Figure 2.

Abdominal radiographs indicating bubbly and linear pneumatosis

Figure 3.

Microscopic and gross pathology delineating pneumatosis in NEC

Figure 4.

Radiolucencies indicating portal venous gas

Pneumatosis is found at least transiently in 70-80% of infants with NEC but may not ever be detected, especially in infants who have received no or minimal enteral feedings or in infants who are very premature (less than 26 weeks postmenstrual age). Portal venous gas was once thought to be an especially ominous finding but more recent studies have debunked this teaching.

Pneumoperitoneum (free air in the abdomen) indicates full thickness intestinal necrosis at one or more locations with escape of intestinal gas into the peritoneal cavity (See Figure 5). A large quantity of free air may be immediately visualized on the KUB. Findings may include marked hepatic hypodensity and perhaps a classic “football sign” caused by an air outline of the falciform ligament. Smaller volumes of free air may be diagnosed on decubitus or cross-table lateral films. A thin air shadow over the liver on a left lateral decubitus may be the only finding to diagnose a pneumoperitoneum (see a large accumulation in Figure 5).

Figure 5.

Radiographs illustrating radiolucent pneumoperitoneum

It is important to position the baby on the left side for a minimum of 5 minutes to allow the air to rise and hence facilitate detection on a left lateral decubitus study. Intestinal preforation is often documented at laparotomy in infants whose serial radiographs never identified free air.

Fixed (adynamic) dilated loops of bowel (“sausage loops”) may include the presence of necrotic intestinal segments (See Figure 6).

Figure 6.

Adynamic loops in serial radiographs indicate possible intestinal necrosis

Bowel wall thickening may be seen consistent with intestinal inflammation and bowel edema.

A “gasless” abdomen or air-fluid levels within loops of bowel may indicate third-spacing within the intestine.

The presence of ascites as suggested by medial displacement of bowel loops on a KUB is generally a late sign (See Figure 6). Ascites generally indicates peritonitis due to occult intestinal perforation or translocation of organisms across a necrotic bowel.

Abdominal ultrasound has been used in some centers to assess the viability of the bowel, to diagnose ascites, to screen for loculated abscesses, to identify pneumatosis and portal venous gas, and to evaluate for volvulus by assessing the relationship of the superior mesenteric artery to this vein. Ultrasound is perhaps a better modality to identify pneumatosis and portal venous gas than plain radiographs, but its utility in the diagnosis of NEC has been limited due to a requirement for technically advanced portable ultrasound machines and highly experienced pediatric ultrasonologists and ultrasonographers.

Contrast studies are performed (upper gastrointestinal series followed by an enema) when an anatomic intestinal obstruction appears to be a more likely diagnosis than NEC.

Abdominal computed tomography (CT) scan has no practical role in the diagnosis of NEC.

Confirming the diagnosis

Refer to the attached algorithm for the diagnosis and treatment of NEC (See Figure 7).

Figure 7.n


A diagnosis of NEC should be suspected when an infant presents with typical gastrointestinal signs (abdominal distention, bloody stools, excessive gastric residuals), systemic signs of sepsis, and laboratory abnormalities.

Findings of pneumatosis intestinalis and/or portal venous gas on abdominal radiographs essentially confirm a diagnosis of NEC. In the absence of these pathognomonic radiographic observations, abnormal CBC findings and unexplained hyponatremia (<130 mEq/L), and clinical ileus in combination with fixed (adynamic) or dilated bowel loops or ascites, can allow a presumptive diagnosis of NEC. Occasionally low molecular weight gastrointestinal contrast studies assist in the confirmation of a diagnosis.

If you are able to confirm that the patient has necrotizing enterocolitis, what treatment should be initiated?

Refer to the algorithm in Figure 7 for the diagnosis and treatment of NE.

When an infant is suspected to have NEC, all enteral feedings and medications should be stopped and parenteral nutrition should be initiated. It is especially important to have a high index of suspicion in the most premature infants because they are less likely to manifest the pathognomomic radiographic findings of NEC.

By the time an infant develops abdominal tenderness or an abdominal mass, NEC has usually progressed to a moderate or advanced stage. In more mature infants, a KUB may be instrumental in establishing a prompt diagnosis. Following culture of blood, urine, and stool, broad spectrum antibiotics should be administered. Regimens include some combination of ampicillin (or cephalosporin), gentamicin ( or other aminoglycoside), and possibly an antibiotic that provides anaerobic coverage (e.g., clindamycin or flagyl) especially if there is bowel perforation or suspected peritonitis.

Intravascular volume status must undergo careful continuous assessment and repletion should be accomplished with crystalloids (e.g., Ringer’s lactate) due to large third spacing losses. Hematocrit should be monitored and packed red cells provided to replace losses due to intestinal bleeding. Pressor support is indicated if sepsis causes loss of vasomotor tone. Adjunctive treatment includes support of pulmonary function with oxygen and appropriate ventilator treatment as well as judicious correction of significant thrombocytopenia and coagulopathy. Co-management with pediatric surgeons is recommended when diagnosis of NEC is strongly suspected or confirmed.

Abdominal girth should be monitored frequently; a sudden increase in abdominal girth warrants an immediate abdominal radiograph to assess for a pneumoperitoneum. Surgery is indicated when a pneumoperitoneum is confirmed or when the totality of clinical evidence suggests intestinal necrosis. Traditional surgical procedures include the excision of clearly necrotic bowel, using multiple resections if needed, to preserve bowel length followed by exteriorzation of viable ends of bowel as stomas with the distal ends exteriorized as a mucous fistula or as a Hartmann’s pouch. In very tiny, critical, or unstable infants, a peritoneal drain can be considered when a focal perforation but not extensive intestinal necrosis is suspected.

In cases where a peritoneal drain is placed, “rescue” laparotomy is indicated should the infant not demonstrate clinical improvement. NEC may occur in any section of the bowel but most commonly affects the terminal ileum and the right colon.

What are the adverse effects associated with each treatment option?

With respect to surgical interventions, insertion of a peritoneal drain can be done at the infant’s bedside and may result in rapid clinical improvement in infants with abdominal distention that is significant enough to cause intraabdominal “compartment syndrome”.

If there is a significant amount of gut necrosis, a peritoneal drain is only a temporizing procedure. Infants may ultimately succumb after a prolonged deterioration if this is the only remedy sought. Hence, “rescue” laparotomy with resection of necrotic bowel and drainage of loculated abscesses is essential for infant survival. However, open laparotomy with intestinal resection is a very invasive procedure and has been associated with greater risks of short bowel syndrome, prolonged hyperalimentation, intestinal stricture, multiple surgeries, and longer lengths of stay.

What are the possible outcomes of necrotizing enterocolitis?

The mortality of NEC in large series is approximately 20-30% with mortality inversely related to postmenstrual age and higher in the sicker infants who have undergone surgical procedures. Perhaps 40% of infants require one or more surgical procedures. Infants who require surgery are also more likely to require longer hospital stays. The two major complications of advanced NEC are stricture formation and short gut syndrome. The average cost of care over 5 years for an NEC survival with short gut syndrome has been estimated to be $1.5 million.

What causes this disease and how frequent is it?

Over the past several decades, the incidence of NEC has not significantly decreased, perhaps because a greater number of very preterm infants are surviving as neonatal intensive care has improved. NEC occurs in 1-5% of all neonatal intensive care admissions and in 3-10% of all very low birth weight infants (birth weight <1500 grams).

NEC remains an important cause of mortality and morbidity in premature infants. Both the incidence and mortality of NEC are inversely related to postmenstrual age. The age of onset of NEC is also inversely related to the postmenstrual age, with onset later in the most preterm infants. For all infants with birth weight < 1500 grams, the average age of onset is 2-3 weeks after birth.

Most cases of NEC are sporadic and have no reproducible seasonal distribution. However, outbreaks of NEC have occurred as clusters of cases in NICUs. Such endemics suggest an infectious etiology, and many have been associated with a common pathologic agent. A number of different bacterial and viral pathogens have been linked to these outbreaks, so that there is no single infectious agent that can be implicated in a common etiologic pathway.

Clinically, NEC must be managed as a local (gastrointestinal) and systemic infectious inflammatory disease. Any infant who develops signs of NEC should be immediately cohorted and cared for using proper isolation precautions.

Genetic factors: Twin studies have suggested a genetic basis for susceptibility to NEC but as yet no specific genetic factors have consistently been linked to a higher risk for NEC.

What complications might you expect from this disease or its treatment?

The two most common significant complications are stricture formation and short gut syndrome.

Anatomic strictures develop in segments of necrotic bowel that undergo sufficient fibrosis during the healing process so that critical narrowing of the intestinal lumen causes signs of obstruction. Strictures occur in 25-40% of all infants with NEC and are most common in infants who do not require surgical intervention.

Symptoms of obstruction (feeding intolerance, abdominal distention, persistent bowel loop dilatation on radiographs, recurrent bloody stools) become apparent upon reintroduction of feedings as early as a few weeks but may not be diagnosed until as late as 6-8 weeks after the acute disease. Functional obstruction can also occur in areas of the bowel that become atonic or remain significantly dilated after resolution of the acute inflammatory process. Diagnosis is made by gastrointestinal contrast studies and surgical treatment is usually required.

Short gut syndrome results when the total length of resected small and large bowel reduces the surface area of the remaining viable bowel so significantly that its digestive and absorptive capacities cannot support enteral delivery of the full caloric requirement for normal growth. Typically infants will appear to tolerate up to a certain daily volume or caloric load. Further increases in volume or nutrient substrate will result in abdominal distention, gastric residuals, emesis, or “dumping” due to rapid gastrointestinal transit. These infants will require prolonged intravenous parenteral nutrition to supplement enteral feedings. Some of these infants may eventually experience sufficient intestinal growth over a couple of years to transition to fully enteral nutrition, while others may require intestinal transplantation.

Other complications include intraabdominal and liver abscesses, development of enterocutaneous fistulae, prolonged cholestasis, hepatic failure, and recurrent blood stream infections associated with the prolonged use of central venous lines.

Are additional laboratory studies available to assist in the diagnosis; even some that are not widely available?


How can necrotizing enterocolitis be prevented?

Despite a plethora of tested interventions, effective and generalizable strategies to prevent NEC remain elusive. Efforts to prevent NEC must logically target one or multiple etiologic factors that are presumed to be of importance: immaturity of the absorptive, digestive, and immunologic functions of the gastrointestinal tract; excessive gastrointestinal inflammatory responses; and colonization of the intestinal lumen with an abnormal microbiota.

It does appear that antenatal (but not postnatal) exposure to corticosteroids confers some protection against NEC, presumably through mediation of key maturational events in the intestinal mucosa. Institution of early, trophic feedings may also prevent the arrest or regression of intestinal maturation.

Whereas systemic intravenous immunoglobulin therapy has not reproducibly decreased the incidence of NEC, one controlled study that has not been reproduced reported that oral supplementation of preterm infants with IgA did reduce the incidence of NEC.

Feeding an infant its own mother’s milk provides several immunological benefits against intestinal pathogens as well as important intestinal growth factors. Such maternal milk feedings have been associated with a decreased risk of NEC. Studies from Hong Kong demonstrated a marked reduction in NEC after institution of very strict precautions that were intended to significantly reduce exposure of infants to enteric pathogens.

Finally, supplementation of oral feedings with a variety of probiotics has shown some promise in preventing NEC, presumably by maintaining or reestablishing a more “normal” intestinal microbiota. However, the long term implications of administering probiotics to premature infants are not clear. Some investigators have raised concern about the potential infectious threat that probiotics pose to neonates.

What is the evidence?

Neu, J, Walker, WA. “Necrotizing enterocolitis”. N Engl J Med. vol. 364. 2011. pp. 255-64. (This recent paper provides a cogent summary of current thoughts about the etiology and prevention of NEC.)

Lin, PW, Nasr, TR, Stoll, BJ. “Necrotizing enterocolitis: recent scientific advances in pathophysiology and prevention”. Semin Perinatol. vol. 32. 2008. pp. 70-82.. (This review provides an excellent summary of the pathophysiology of NEC and of prevention strategies, as well as a state-of-the-art discussion of the evidence that speaks to genetic predisposition.)

Lin, PW, Stoll, BJ. “Necrotizing enterocolitis”. Lancet. vol. 368. 2006. pp. 1271-83. (Another good review of NEC.)

Afrazi, A, Sodhi, CP, Richardson, W. “New insights into the pathogenesis and treatment of necrotizing enterocolitis: Toll-like receptors and beyond”. Pediatr Res. vol. 69. 2011. pp. 183-8. (This review summarizes how activation of Toll-like receptors by intestinal bacteria signals an excessive inflammatory response, and provides a rationale for research into the prevention of NEC via silencing of Toll-like receptors in the intestinal epithelium.)

Morowitz, MJ, Poroyko, V, Caplan, M. “Redefining the role of intestinal microbes in the pathogenesis of necrotizing enterocolitis”. Pediatrics. vol. 125. 2010. pp. 777-85. (This paper argues that the nature of the intestinal microbiota is an important risk factor for NEC.)

Claud, EC, Walker, WA. “Hypothesis: inappropriate colonization of the premature intestine can cause necrotizing enterocolitis”. FASEB J. vol. 15. 2001. pp. 1398-1403. (Good early review of potential role of intestinal microbes in the pathogenesis of NEC.)

Alfaleh, K, Anabrees, J, Bassler, D, Al-Kharfi, T. “Probiotics for prevention of necrotizing enterocolitis in preterm infants”. Cochrane Database Syst Rev. vol. 3. 2011. pp. CD005496(Recent comprehensive review of probiotics and NEC.)

Indiro, F, Neu, J. “The intestinal microbiome of infants and the use of probiotics”. Curr Opin Pedatr. vol. 23. 2011. pp. 145-50. (Some cautionary thoughts about using available probiotics routinely in premature infants.)

Fanaroff, AA, Stoll, BJ, Wright, LL. “NICHD Neonatal Research Network. Trends in neonatal morbidity and mortality for very low birthweight infants”. Am J Obstet Gynecol. vol. 196. 2007. pp. 147.e1-8. (This prospective prevalence study by the NICHD network demonstrated significant intercenter variability in birth weight specific incidences of NEC but among all centers detected no cumulative change in incidence between 1997 and 2002.)

Sharma, F, Tepas, JJ, Hudak, ML. “Portal venous gas and surgical outcome of neonatal necrotizing enterocolitis”. J Pediatr Surg. vol. 40. 2005. pp. 371-6. (This prospective prevalence study describes the experience of a single center over 12 years which substantiated that severity of NEC, and not the presence of portal venous gas per se, correlated with mortality.)

Nanthakumar, N, Meng, D, Goldstein, AM. “The mechanism of excessive intestinal inflammation in necrotizing enterocolitis: an immature innate immune response”. PLoS One. vol. 6. 2011. pp. e17776(This case controlled molecular study of resected human intestine concludes that the excessive inflammatory response of the immature intestine, a key factor in the etiology of NEC, results from developmental immaturity in immune response genes.)

Sharma, R, Hudak, ML, Tepas, JJ. “Impact of gestational age on the clinical presentation and surgical outcome of necrotizing enterocolitis”. J Perinatology. vol. 26. 2006. pp. 342-7. (Data presented in this prospective observational single center prevalence study convincingly demonstrates that the signs of NEC in the preterm infant change with increasing maturity.)

Shankaran, K, Puckett, B, Lee, DS. “Variations in incidence of necrotizing enterocolitis in Canadian neonatal intensive care units”. J Pediatr Gastroenterol Nutr. vol. 39. 2004. pp. 366-72. (Report of a prospective, observational multicenter study of NEC in the Canadian Neonatal Network.)

Kuppala, VS, Meinzen-Derr, J, Morrow, AL, Schibler, KR. “Prolonged initial empirical antibiotic treatment is associated with adverse outcomes in premature infants”. J Perinatol. 2011. (Case control study that suggests that antibiotic usage is a risk factor for NEC.)

Sharma, R, Tepas, JJ, Hudak, ML. “Neonatal gut injury and infection rate: impact of surgical debridement on outcome”. Pediatr Surg Int. vol. 21. 2005. pp. 977-82. (This prospective single center prevalence study concluded that infants with severe neonatal gut injury [NEC and IIP] were more likely to survive and to incur infectious morbidity if they underwent surgical debridement.)

Sharma, R, Tepas, JJ, Mollitt, DL. “Surgical management of bowel perforations and outcome in very low-birth-weight infants (< or = 1,200 g)”. J Pediatr Surg. vol. 39. 2004. pp. 190-4. (This study argues for early laparotomic intervention in infants with bowel perforations and serious systemic illness.)

Moss, RL, Dimmitt, RA, Barnhart, DC. “Laparotomy versus peritoneal drainage for necrotizing enterocolitis and perforation”. N Engl J Med. vol. 354. 2006. pp. 2225-34. (A randomized controlled trial of two surgical treatments.)

Blakely, ML, Tyson, JE, Lally, KP. “Laparotomy versus peritoneal drainage for necrotizing enterocolitis or isolated intestinal perforation in extremely low birth weight infants: outcomes through 18 months adjusted age”. Pediatrics. vol. 117. 2006. pp. e680-7. (A prospective multicenter cohort study by the NICHD neonatal network that comes out in favor of laparotomy.)

Rees, CM, Eaton, S, Kiely, EM. “Peritoneal drainage or laparotomy for neonatal bowel perforation? A randomized controlled trial”. Ann Surg. vol. 248. 2008. pp. 44-51. (Another study that demonstrates that many infants treated by peritoneal drainage later require laparotomy.)

Sola, JE, Tepas, JJ, Koniaris, LG. “Peritoneal drainage versus laparotomy for necrotizing enterocolitis and intestinal perforation: a meta-analysis”. J Surg Res. vol. 161. 2010. pp. 95-100. (A recent meta-analysis of five studies that finds peritoneal drainage to be associated with higher mortality.)

Millar, M, Wilks, M, Fleming, P, Costeloe, K. “Should the use of probiotics in the preterm be routine?”. Arch Dis Chld Fetal Neonatal Ed. 2010. (This editorial summarizes the pros and cons associated with adopting probiotics as a standard preventive treatment for NEC. In the end, the authors argue that additional information from controlled studies that use a well-defined and regulated probiotic is needed.)

Foster, J, Cole, M. “Oral immunoglobulin for preventing necrotizing enterocolitis in preterm and low birth-weight neonates”. Cochrane Database Syst Rev. vol. 1. 2004. pp. CD001816(Review of this issue.)

Meinzen-Derr, J, Poindexter, B, Wrage, L. “Role of human milk in extremely low birth weight infants' risk of necrotizing enterocolitis or death”. J Perinatol. vol. 29. 2009. pp. 57-62. (One of several studies to suggest that human milk feedings reduces the likelihood of NEC.)

Lin, HC, Su, BH, Chen, AC. “Oral probiotics reduce the incidence and severity of necrotizing enterocolitis in very low birth weight infants”. Pediatrics. vol. 115. 2005. pp. 1-4. (Early randomized controlled trial of probiotics and NEC.)

Manzoni, P, Lista, G, Gallo, E. “Routine administration in VLBW infants: a retrospective, 6-year cohort study”. Early Hum Dev. vol. 87. 2011. pp. S35-8. (This study provides evidence for long-term safety of a single bacterial probiotic in preterm infants.)

Ongoing controversies regarding the etiology, diagnosis, treatment

Surgical treatment: peritoneal drain versus laparotomy

A surgeon may choose either to place a peritoneal drain or to perform a laparotomy as an initial procedure in infants with severe NEC. In 2001, a meta-analysis of 10 published investigations could not determine which procedure was the better choice to improve survival and/or to minimize morbidity. Subsequently, a large prospective single center study showed that infants with severe NEC with perforation who were treated initially with laparotomy fared better than infants treated with a peritoneal drain.

Insertion of a drain appeared to be a reasonable choice for infants with focal perforation who did not demonstrate significant systemic signs. These investigators suggested that the severity of the underlying illness and not the birth weight should determine the procedure. In this study, a “rescue laparotomy” was done if treatment with peritoneal drainage failed.

A multicenter prospective cohort study by the National Institute of Child Health and Human Development Neonatal Research Network found a lower survival rate and an increased incidence of neurodevelopmental impairment among infants treated with a peritoneal drain compared to infants treated with laparotomy. Another international multicenter randomized controlled trial evaluated 69 infants and found that 74% of infants treated initially with peritoneal drainage later required laparotomy. This study argued that initial placement of a peritoneal drain was ineffective as either a temporizing measure or as a definitive treatment and recommended that a timely “rescue laparotomy” should be considered in infants treated by initial peritoneal drainage.

A recent meta-analysis of five prospective studies conducted during 2000-2008 also reported increased mortality with peritoneal drainage. Nonetheless, none of these studies critically compared the choice of surgical procedure in infants with presumed focal perforation without advanced signs of systemic disease.

Probiotics for the prevention of NEC

Several clinical trials and meta-analyses have demonstrated that prophylactic treatment with probiotics decreases the likelihood of NEC and mortality associated with NEC. Yet, there is still controversy about using probiotics in premature infants. One concern is that administration of live bacteria to very low birth weight infants may increase the risk of systemic infection. It is also uncertain how probiotics will impact the long-term health of an infant’s immune system.

The initial neonatal period is a critical time during which an infant begins to develop a healthy state of immune-tolerance. Interactions of luminal microbes with the intestinal mucosa and immune receptors may establish a healthy immune tolerance or may trigger a trajectory toward development of auto-immune or allergic diseases later in life. Reassuringly, results of a 6 year follow-up of 743 infants in one clinical study of the probiotic Lactobacillus rhamnosus GG (LGG) provided strong evidence of long-term safety.

One of the problems with generalizing from available clinical studies is that different species of probiotics have been used in trials characterized by varying dosing strategies and timing of administration. A further difficulty is that there has not been adequate quality assurance, standardization, or regulation of the probiotic products. As currently marketed, it is not certain that a probiotic is active (live) during administration in the intensive care nursery. Nonetheless, probiotics represent a sufficiently promising approach for the prevention of NEC.

Further large double-blind randomized controlled trials are needed to determine the optimal preparation, dosing, and duration of probiotic treatment. Adjunct studies that will correlate outcomes with patterns of intestinal microbiota would assist in evaluating the potential benefits of probiotics.