Enterocolitis Necrotizante 2018

Seminars in Fetal and Neonatal Medicine 23 (2018) 369
Contents lists available at ScienceDirect
Seminars in Fetal and Neonatal Medicine

journal homepage: www.elsevier.com/locate/siny
Editorial
Necrotizing enterocolitis T
Necrotizing enterocolitis (NEC) has emerged as a major cause of microcirculation, and the developing microbiota, which appears to
mortality and morbidity in infants. Despite considerable research, become “dysbiotic” prior to clinical presentation of this disease. How
prevention and/or cure remain elusive because NEC is not a clear this interplay sets up a “perfect storm” scenario that culminates in NEC
target. It represents more than one disease with multifactorial patho- is also discussed.
genic mechanisms. Databases for NEC are blurry at best; hence, epi- How to care for infants who have been diagnosed with NEC remains
demiologic research on this disease is hampered by the lack of a clear controversial. Among the most misunderstood issues in care of at-risk
definition of what constitutes the major form of this disease. Likewise, babies is the role of feeding in the pathogenesis of NEC. Feeding volume
animal models that putatively represent the disease are off mark since and composition are widely thought to be reasons for NEC development
they only represent a limited component of the disease seen in human in certain infants. The role of feeding and composition is critically
infants. Finding genomic, proteomic, inflammatory mediator and other discussed. More recent studies focusing on surgical care of infants with
“omics”-based predictive and diagnostic biomarkers will be difficult NEC are also summarized.
until the different subsets of what we are calling this disease are better It has recently become clear that the morbidity seen in infants who
delineated. Recognition and diagnosis of NEC currently depend largely develop NEC is not only that seen in the neonatal intensive care unit
on a diffuse set of non-specific clinical, radiologic and laboratory eva- (NICU), with increased need for antibiotics, parenteral nutrition, pro-
luations. longed hospitalization and short gut. NEC also presents a major fi-
In this issue of Seminars in Fetal and Neonatal Medicine, we examine nancial burden, which is also summarized. Furthermore, the morbidity
several aspects of NEC. Studies using epidemiologic approaches to of NEC extends well beyond the NICU and the infant gastrointestinal
evaluate what is currently known about the risk factors for develop- tract. This disease has also been found to be associated with major
ment of NEC will be critically analyzed. The search for better diagnostic neurodevelopmental delays, which will also be discussed.
and predictive biomarkers that will help us better delineate this disease The guest editor would like to acknowledge that the reviews pre-
is summarized. Several aspects of disease pathogenesis include genetic sented in this volume provide only a part of an overview of this disease,
predisposition, the role of the interplay between the developing mi- but which nevertheless point to several areas where intensive research
crobiome and innate and adaptive immunity, and the developing mi- based on a clearer definition of this disease will lead to its eradication in
crocirculation. the near future.
Suspicions that the pathogenesis of NEC may have a genetic pre-
disposition come from epidemiologic studies. Newly developed DNA- Josef Neu
based technologies used in finding genetic loci related to NEC are de- Department of Pediatrics, University of Florida, Gainesville, FL, USA
scribed. Other components of the pathophysiology of this disease in- E-mail address: neuj@peds.ufl.edu
clude the interplay between host immunity, the developing
https://doi.org/10.1016/j.siny.2018.08.009
1744-165X/ © 2018 Elsevier Ltd. All rights reserved.
Descargado para Anonymous User (n/a) en Hospital de Pediatria de ClinicalKey.es por Elsevier en diciembre 16, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
Seminars in Fetal and Neonatal Medicine 23 (2018) 370–373

Necrotizing enterocolitis comes in different forms: Historical perspectives T

and defining the disease
J. Neua,∗, N. Modib, M. Caplanc
a
University of Florida, College of Medicine, Gainesville, FL, USA
b
Imperial College London, London, UK
c
University of Chicago Pritzker School of Medicine, NorthShore University Health System, Evanston, IL, USA
A R T I C LE I N FO A B S T R A C T
Keywords: The specific cause of what is commonly referred to as necrotizing enterocolitis (NEC) disease has been elusive
Necrotizing enterocolitis largely because it is becoming clear that this entity represents more than one disease with multifactorial pa-
Pathophysiology thogenic mechanisms. Furthermore, finding clear and consistent diagnostic biomarkers will be difficult until the
Differential diagnosis different subsets of what we are calling this disease are better delineated. In this introductory chapter, we discuss
different disease entities that are frequently termed “NEC” in the newborn infant. We hope this will set the stage
for more focused research and development of preventative measures for at least the most common forms of this
disease.
1. Historical perspectives diagnostic biomarkers that are both highly sensitive and specific for the
various categories of NEC is one of the major challenges facing the field.
Neonatal intensive care is a relatively new field. In the USA, the If in fact “NEC” represents different diseases, we need different bio-
organization of intensive care for preterms, newborns, and young in- markers to differentiate these subsets. Once this is accomplished,
fants began in the 1960s. In the UK, the establishment of neonatology as treatment and preventative measures would differ depending on the
an approved pediatric training sub-specialty took place in the 1980s. individual pathophysiology.
This provision of an organizational structure for care of the critically ill In this introductory chapter, we wish to initiate a delineation of
newborn placed major emphasis on improvement of respiratory sup- different disease entities that are frequently termed “NEC” in the
port, management of fluids, electrolytes and glucose, thermoregulation, newborn infant (Box 1).
and infections especially with staphylococcal disease. Necrotizing en-
terocolitis (NEC) was still an uncommon entity. Reports of NEC-like 2. Stage 1 “NEC″
illnesses in newborns date back to the early 1800s [1,2], but no mention
is made of NEC in several textbooks of newborn medicine from the Because of the potential rapid progression of NEC, it is highly de-
1960s. Since then, NEC has emerged as a major cause of mortality and sirable to diagnose the condition at its earliest stages. In the late 1970s,
morbidity in infants, especially those born preterm [3–6]. Dr Martin Bell proposed clinical staging criteria (now known as Bell's
The specific cause of the disease remains enigmatic [7]. Despite stages of NEC) [13]. Despite modifications made to this system over
considerable research about NEC, we have not yet found the means to several decades, it has become outdated. Increased survival of ex-
diagnose this disease reliably using consistent criteria, nor to treat it tremely immature infants, the majority of whom exhibit “Stage 1” signs
effectively when it occurs. The pathophysiology of NEC is considered and symptoms without actually developing confirmed intestinal ne-
“multifactorial” [7,8]. This is not surprising because what we term crosis, causes confusion. Furthermore, there is no evidence to suggest
“NEC” appears to represent a range of pathophysiology, similar to that that stage 1 NEC progresses towards stage 2 or stage 3, and in fact these
for diabetes [9,10]. It is clear that we do not have a precise definition cases of feeding intolerance are highly non-specific, and when recorded
for NEC [11,12]. Can we even aim to achieve this if NEC represents as “NEC” make subsequent interpretation prone to inaccuracies. It is
more than one disease? A single definition for NEC (as with diabetes) becoming clear that deciding when to continue with medical inter-
would need to take into account a range of gastrointestinal morbidities vention or proceed to surgery is more important than formal staging.
with differing pathophysiologies. The lack of clear and consistent However, staging may be useful when attempting to evaluate the
∗
Corresponding author. University of Florida, College of Medicine, Gainesville, FL, USA.
E-mail address: neuj@peds.ufl.edu (J. Neu).
1744-165X/ © 2018 Published by Elsevier Ltd.
J. Neu et al. Seminars in Fetal and Neonatal Medicine 23 (2018) 370–373
Box 1
Some imposters of necrotizing enterocolitis (NEC)
“Stage 1, 2 and 3” NEC

Cardiogenic ischemia
Variants of food protein-induced enterocolitis syndrome
Spontaneous isolated intestinal perforations
Misrepresentation of pneumatosis – “poopatosis”
Placement of drain for pneumoperitoneum without direct visualization of bowel
Congenital bowel anomalies (e.g. Hirschsprung's aganglionosis)
impact of an intervention or treatment in clinical trials or epidemiologic plausible that compromised mesenteric blood flow also contributes to
studies. the pathophysiology of preterm NEC. However, ischemic bowel due to
In the past several years, research networks such as CHD may benefit from strict control of cardiac output and enhancement
Vermont–Oxford, National Institutes of Health Neonatal Research of splanchnic perfusion. Although the authors suggested that this dis-
Network (NRN) and Pediatrix have begun to limit the criteria for NEC ease entity be termed cardiogenic NEC (“CNEC”), this term may still
diagnosis to Bell's stages 2 and 3. This remains problematic. Stage 2 cause confusion with the more typical form of NEC and thus we propose
relies on clinical signs and radiographic criteria such as gas in the bowel to name this disease “cardiogenic ischemic necrosis of the intestine” to
wall (pneumatosis intestinalis) or portal venous gas. The Duke clearly differentiate it from the “classic” forms.
Assessment scale has attempted to provide a broader range of radio-
graphic criteria, which include persistent loops of bowel on sequential
radiographs, intestinal dilatation and separation of bowel loops, all of 4. NEC related to feed composition
which can be highly non-specific [14,15]. Pneumatosis recognition can
be challenging. A bubbly appearance may be due to stool in the bowel It is becoming increasingly clear that many babies who develop a
lumen and misconstrued as representing gas in the bowel wall. syndrome with some typical signs of NEC may actually have a distinct
If pneumoperitoneum is found, this could be interpreted as stage 3 disease related to the composition of feeds. Powel [23] first described
(surgical) NEC, but if the surgeon does not visualize the bowel at la- this in 1976 in two low-birth-weight infants who developed a syndrome
parotomy, and instead places a peritoneal drain [16], whether the of vomiting, distension, septic appearance, and bloody diarrhea. This
pneumoperitoneum resulted from NEC, spontaneous intestinal per- was seen after ingestion of cow milk-based formula initially, and, later,
foration or other cause may never be definitively determined [17]. soy-based formulas. These symptoms resolved with intravenous fluids
Nevertheless, in most previous studies, cases of pneumoperitoneum in and alimentation.
which peritoneal drains were placed were designated as being “NEC”, Numerous reports have identified infants outside of the neonatal
leading to database inaccuracies. period with bloody stools, abdominal distension, flank discoloration
and pneumatosis intestinalis that improved after cow's milk protein was
excluded but relapsed after challenge [23–25]. The diagnosis of NEC
3. Ischemic bowel due to congenital heart disease was made in some and others were diagnosed as milk-induced en-
terocolitis. This raises the question of whether food protein-induced
Congenital heart disease is a risk factor for the development of NEC- enterocolitis syndrome (FPIES) occurs in preterm neonates [26]. If so,
like clinical presentation seen predominantly in full-term neonates could this cause diagnostic confusion with classic NEC?
[18–21]. The risk of “NEC” in a population of infants with congenital Food protein-induced enterocolitis (FPIES) is a non-im-
heart disease (CHD) compared with the entire preterm and term neo- munoglobulin E-mediated syndrome resulting in hypersensitivity to
natal population is nearly 10–100-fold [20]. Univentricular heart dis- food antigens. This entity is primarily described in older, not preterm
ease, especially hypoplastic left heart syndrome and univentricular infants [27]. Patients with FPIES typically present in the first months
anomalies with arch obstruction, are most highly associated with NEC. after birth with vomiting, diarrhea, hematochezia, or lethargy within
Truncus arteriosus and aortopulmonary window also accounted for one to four weeks after initial exposure to a triggering antigen. Similar
several babies with “NEC”. This suggested that decreased mesenteric to NEC, FPIES is more common in infants fed formula rather than
blood flow is an important factor in the pathophysiology of NEC in human milk. As with NEC, the presence of transforming growth factor-β
neonates with these forms of CHD. Common features of many of the and IgA in breastmilk may be protective for infants who are exclusively
structural cardiac lesions associated with NEC are widened pulse breastfed, but nonetheless FPIES may still occur in breastfed infants
pressure and low diastolic pressure. This abnormal physiology often [28,29].
results in retrograde diastolic flow in the descending aorta [22], leading A clear diagnosis of FPIES is difficult. Laboratory findings cannot
to mesenteric ischemia. confirm the diagnosis of FPIES alone, but general trends are common in
In an evaluation of outcomes of patients with congenital heart dis- these patients. These include leukocytosis with an eosinophil pre-
ease-related NEC compared to preterm neonatal NEC, better short- and dominance, thrombocytosis, hemoccult positive stools, anemia, hy-
long-term outcomes were found for CHD-related NEC [21]. The authors poalbuminemia, and, in some cases, radiographic findings of pneuma-
concluded that the “better outcomes, in conjunction with previous tosis intestinalis [30,31]. An oral food challenge is the “references
studies supporting the role of decreased mesenteric flow in NEC pa- standard” diagnostic criterion for FPIES.
thogenesis, suggest that the constellation of signs and symptoms labeled It is thus important to discern FPIES from classic NEC, as treatment
“NEC” in the cardiac patient should be distinguished from traditional is different. With NEC, the patient is given antibiotics, made “nil by
NEC.” They also concluded that there should be a change in nomen- mouth”, and often requires parenteral nutrition for extended periods. In
clature to distinguish these entities. Whether such a change would contrast, management of FPIES is to remove the offending antigen with
improve the understanding of the disease processes and impact in- dietary elimination, and typically intestinal necrosis does not occur as it
dividualized management is unclear; nonetheless it may improve our does frequently in NEC. Typically, breastfeeding can be continued with
prognostication for parents in these unique situations. It is also highly maternal dietary modifications (as breastmilk is generally an
371
uncommon trigger). If the baby is formula-fed, a switch to a hy- 6. Transfusion-related NEC

poallergenic regimen, such as an extensively hydrolyzed or an amino
acid-based formula, alleviates the symptoms. With additional studies One of the most common practices in neonatal intensive care is
that provide more sensitive and specific biomarkers that differentiate transfusion of packed red blood cells. Over the past couple of decades,
FPIES from other forms of “NEC”, it may ultimately be possible to more several observational studies have demonstrated a temporal association
clearly differentiate between these entities and further refine our between packed red blood cell transfusions and NEC [45]. This has led
treatment regimens. to the belief that there exists a distinct entity that has been termed
“transfusion-associated NEC” (TANEC) [46]. Because of this risk, many
neonatologists have recommended withholding feedings for variable
5. Spontaneous intestinal perforation periods of time before, during, and after packed red blood cell trans-
fusions. More recently, it has been questioned whether there is a causal
Spontaneous intestinal perforation (SIP) was first reported as a relationship between transfusions and NEC, and several randomized
distinct diagnosis from NEC in the late 1980s and early 1990s [32,33]. controlled trials do not appear to support such a causal relationship
SIP mainly affects extremely low-birth-weight (ELBW) infants at an [47].
early postnatal age. SIP is differentiated from NEC and characterized as A recent critical analysis of the previous studies that relate packed
an isolated perforation without surrounding necrosis or neutrophil in- red blood cell transfusions to NEC utilized the “Grading of
filtrate. The perforation is often accompanied by a focal thinning or Recommendations Assessment, Development and Evaluation” (GRADE)
absence of the intestinal muscularis propria [34]. The pathophysiology methodology [48] to summarize the quality of existing evidence [49].
of SIP is poorly understood. Onset typically occurs earlier than NEC, in This review yielded “low” to “very low” confidence for a true associa-
the first two weeks after birth, in ELBW infants. Studies to date have not tion. The observational and randomized studies exhibited pitfalls in-
clearly delineated whether postnatal corticosteroids and/or non-ster- cluding small sample size, lack of accounting for baseline hematocrit
oidal anti-inflammatory agents are associated with SIP [35,36]. Infants prior to transfusions, and a lack of clarity in defining NEC. Furthermore,
with SIP are often relatively stable in the early stages, without systemic the observational versus the randomized studies did not provide con-
signs and symptoms of severe illness. They do not show radiologic sistent results in terms of direction of effects, with the observational
features of pneumatosis intestinalis and portal venous gas as seen in studies suggesting higher NEC with transfusions and the randomized
NEC [37]. Despite these differences, NEC and SIP are often treated si- studies suggesting lower NEC, possibly associated with pre-existing
milarly, with placement of peritoneal drains. Prospective, randomized anemia. In a recent evaluation, Patel et al. [50] demonstrated that it
studies in the USA [38] and UK [39] showed no difference in clinical was severe anemia rather than the packed cell transfusion that con-
outcomes between laparotomy versus peritoneal drain placement for tributed significantly to the development of NEC. Thus, the evidence at
bowel perforation, but in the UK study, 74% of the neonates treated this time is not sufficient to support TANEC as a distinct clinical entity.
with primary peritoneal drainage required delayed laparotomy, and the Interventions such as withholding feedings during transfusions have not
authors suggested that “if a drain is inserted, a timely “rescue” lapar- been adequately studied, though a large randomized prospective trial is
otomy should be considered.” Currently, in many institutions, laparo- in progress.
tomies are done as a second procedure after failure of the drain, but
whether these are done in a “timely” manner that prevents mortality 7. Conclusion
and additional morbidity remains debatable. The choice of initial la-
parotomy or peritoneal drain thus remains controversial. We are If we are to make progress in the prevention and treatment of NEC, a
awaiting the data from a large, prospective randomized trial from the clearer definition based on improved understanding of the pathophy-
NIH Neonatal Research Network from a study designed to determine siology of the different entities (and we have not provided a compre-
whether either of these modalities improves neurodevelopmental im- hensive list) that are called “NEC” is needed. Diagnostic criteria should
pairment associated with surgical NEC. Proponents of initial drain be evidence-based to perform well in respect of sensitivity, specificity
placement hypothesize that it allows stabilization before laparotomy, and predictive value to aid the clinician in differentiating the classic
but a prospective randomized trial (NET Trial Group) showed that form of the disease from other entities such as spontaneous intestinal
peritoneal drainage does not immediately improve clinical status. perforation, primary intestinal ischemia, sepsis, transfusion-associated
Further research is needed to confirm the optimal surgical approach for enterocolitis, enteropathies due to congenital anomalies, and food
these complex patients [40]. protein intolerance. For the moment we suggest that clinicians should
Several investigations have attempted to differentiate SIP from NEC. record each feature individually. These data include patient descriptors
Studies compared circulating inter-alpha inhibitor proteins (IaIp), a (e.g. gestational age, postnatal age, sex, birth weight), co-morbidities
family of structurally related serine protease inhibitors that play a role (e.g. CHD, cytomegalovirus status, transfusion history), X-ray findings
in systemic inflammation. In patients with NEC versus patients with (e.g. pneumatosis, portal venous gas, fixed bowel loop, distended
non-specific disorders, IaIp correlated with disease progression when bowel), clinical findings (e.g. distended, tender, discoloured abdomen,
measured serially [41]. The authors suggested that this test may serve bilious aspirates/vomits, bloody stool), laboratory investigations (e.g.
as a surrogate to monitor response to therapy in NEC. However, they platelet count, C-reactive protein, blood culture) and surgical findings
also evaluated this biomarker and found that, similar to the responses (e.g. pneumatosis, perforation, necrotic bowel, area of involvement).
seen in NEC, it also was lower in babies with sepsis [42]. The specificity This would enable data to be incorporated into systematic reviews and
of this test as a specific biomarker for NEC that differentiates between meta-analyses, and cases to be classified according to predefined cri-
sepsis, NEC and SIP thus is questionable. Other studies employed im- teria [12].
munoregulatory protein profiles of NEC versus SIP in preterm infants,
finding signatures that differed between these diseases [43]. The same Practice points
group demonstrated genome-wide expression differences between in-
testinal tissues after known episodes of NEC or SIP [43] and microRNA • The clinician needs to understand that there are several different
differences between infants with these diagnoses [44]. However, de- entities that represent the disease we call NEC.
spite their importance and promise for our ability to differentiate these • These differ in term of diagnostic criteria and treatment.
two entities, the clinical efficacy of these biomarkers remains to be • A better delineation of these various subsets of NEC will provide for
evaluated by additional studies. more targeted diagnosis, prevention and treatment.
372
Conflicts of interest presenting as recurrent necrotizing enterocolitis in preterm neonates. J Perinatol

2010;30:431–3.
[25] Coviello C, Rodriquez DC, Cecchi S, et al. Different clinical manifestation of cow's
None declared. milk allergy in two preterm twins newborns. J Matern Fetal Neonatal Med
2012;25(Suppl 1):132–3.
Funding sources [26] Murch SH. Cow’s-milk protein as a specific immunological trigger of necrotising
enterocolitis – or food protein-induced enterocolitis syndrome in disguise? J Pediatr
Gastroenterol Nutr 2013;56:3–4.
None. [27] Bingemann TA, Sood P, Jarvinen KM. Food protein-induced enterocolitis syndrome.
Immunol Allergy Clin 2018;38:141–52.
[28] Monti G, Castagno E, Liguori SA, et al. Food protein-induced enterocolitis syndrome
References by cow's milk proteins passed through breast milk. J Allergy Clin Immunol
2011;127:679–80.
[1] Caplan MS, Fanaroff A. Necrotizing: a historical perspective. Semin Perinatol [29] Kaya A, Toyran M, Civelek E, Mısırlıoglu ED, Kırsaçlıoglu CT, Kocabaş CN. Food
2017;41:2–6. protein-induced enterocolitis syndrome in two exclusively breastfed infants. Pediatr
[2] Obladen M. Necrotizing enterocolitis – 150 years of fruitless search for the cause. Allergy Immunol 2016;27:749–50.
Neonatology 2009;96:203–10. [30] Nowak-Wegrzyn A, Muraro A. Food protein-induced enterocolitis syndrome. Curr
[3] Bisquera JA, Cooper TR, Berseth CL. Impact of necrotizing enterocolitis on length of Opin Allergy Clin Immunol 2009;9:371–7.
stay and hospital charges in very low birth weight infants. Pediatrics [31] Mehr S, Kakakios A, Frith K, Kemp AS. Food protein-induced enterocolitis syn-
2002;109:423–8. drome: 16-year experience. Pediatrics 2009;123:e459–64.
[4] Stey A, Barnert ES, Tseng CH, et al. Outcomes and costs of surgical treatments of [32] Aschner JL, Deluga KS, Metlay LA, Emmens RW, Hendricks-Munoz KD.
necrotizing enterocolitis. Pediatrics 2015;135:e1190–7. Spontaneous focal gastrointestinal perforation in very low birth weight infants.
[5] Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011;364:255–64. J Pediatr 1988;113:364–7.
[6] Fullerton BS, Hong CR, Velazco CS, et al. Severe neurodevelopmental disability and [33] Meyer CL, Payne NR, Roback SA. Spontaneous, isolated intestinal perforations in
healthcare needs among survivors of medical and surgical necrotizing enterocolitis: neonates with birth weight less than 1,000 g not associated with necrotizing en-
a prospective cohort study. J Pediatr Surg 2017 Oct 12. pii: S0022-3468(17) terocolitis. J Pediatr Surg 1991;26:714–7.
30651-6. [34] Lai S, Yu W, Wallace L, Sigalet D. Intestinal muscularis propria increases in thick-
[7] Neu J. Necrotizing enterocolitis: the mystery goes on. Neonatology ness with corrected gestational age and is focally attenuated in patients with iso-
2014;106:289–95. lated intestinal perforations. J Pediatr Surg 2014;49:114–9.
[8] Hackam D, Caplan M. Necrotizing enterocolitis: pathophysiology from a historical [35] Stark AR, Carlo WA, Tyson JE, et al. Adverse effects of early dexamethasone
context. Semin Pediatr Surg 2018;27:11–8. treatment in extremely-low-birth-weight infants. N Engl J Med 2001;344:95–101.
[9] Gordon PV, Swanson JR, Attridge JT, Clark R. Emerging trends in acquired neonatal [36] Paquette L, Friedlich P, Ramanathan R, Seri I. Concurrent use of indomethacin and
intestinal disease: is it time to abandon Bell's criteria? J Perinatol 2007;27:661–71. dexamethasone increases the risk of spontaneous intestinal perforation in very low
[10] Torrazza RM, Li N, Neu J. Decoding the enigma of necrotizing enterocolitis in birth weight neonates. J Perinatol 2006;26:486–92.
premature infants. Pathophysiology 2014;21:21–7. [37] Pumberger W, Mayr M, Kohlhauser C, Weninger M. Spontaneous localized in-
[11] Gordon PV, Swanson JR, MacQueen BC, Christensen RD. A critical question for NEC testinal perforation in very-low-birth-weight infants: a distinct clinical entity dif-
researchers: can we create a consensus definition of NEC that facilitates research ferent from necrotizing enterocolitis. J Am Coll Surg 2002;195:796–803.
progress? Semin Perinatol 2017;41:7–14. [38] Moss RL, Dimmitt RA, Barnhart DC, et al. Laparotomy versus peritoneal drainage
[12] Battersby C, Longford N, Costeloe K, Modi N. Development of a gestational age- for necrotizing enterocolitis and perforation. N Engl J Med 2006;354:2225–34.
specific case definition for neonatal necrotizing enterocolitis. JAMA Pediatrics [39] Rees CM, Eaton S, Kiely EM, Wade AM, McHugh K, Pierro A. Peritoneal drainage or
2017;171:256–63. laparotomy for neonatal bowel perforation? A randomized controlled trial. Ann
[13] Bell MJ, Ternberg JL, Feigin RD, et al. Neonatal necrotizing enterocolitis. Surg 2008;248:44–51.
Therapeutic decisions based upon clinical staging. Ann Surg 1978;187:1–7. [40] Rees CM, Eaton S, Khoo AK, Kiely EM, Pierro A. Peritoneal drainage does not sta-
[14] Coursey CA, Hollingsworth CL, Wriston C, Beam C, Rice H, Bisset 3rd G. bilize extremely low birth weight infants with perforated bowel: data from the NET
Radiographic predictors of disease severity in neonates and infants with necrotizing Trial. J Pediatr Surg 2010;45:324–8. discussion 328–29.
enterocolitis. Am J Roentgenol 2009;193:1408–13. [41] Chaaban H, Shin M, Sirya E, Lim YP, Caplan M, Padbury JF. Inter-alpha inhibitor
[15] Markiet K, Szymanska-Dubowik A, Janczewska I, Domazalska-Popadiuk I, protein level in neonates predicts necrotizing enterocolitis. J Pediatr
Zawadzka-Kepczynska A, Bianek-Bodzak A. Agreement and reproducibility of 2010;157:757–61.
radiological signs in NEC using the duke abdominal assessment scale (DAAS). [42] Chaaban H, Singh K, Huang J, Siryaporn E, Lim YP, Padbury JF. The role of inter-
Pediatr Surg Int 2017;33:335–40. alpha inhibitor proteins in the diagnosis of neonatal sepsis. J Pediatr
[16] Rao SC, Basani L, Simmer K, Samnakay N, Deshpande G. Peritoneal drainage versus 2009;154:620–2. e621.
laparotomy as initial surgical treatment for perforated necrotizing enterocolitis or [43] Chan KY, Leung FW, Lam HS, et al. Immunoregulatory protein profiles of necro-
spontaneous intestinal perforation in preterm low birth weight infants. Cochrane tizing enterocolitis versus spontaneous intestinal perforation in preterm infants.
Database Syst Rev 2011;6:CD006182. PLoS One 2012;7:e36977.
[17] Gordon PV. Treatment of necrotizing enterocolitis. N Engl J Med 2006;355:846–7. [44] Ng PC, Chan KY, Leung KT, et al. Comparative MiRNA expressional profiles and
author reply 847. molecular networks in human small bowel tissues of necrotizing enterocolitis and
[18] Leung MP, et al. Necrotizing enterocolitis in neonates with symptomatic congenital spontaneous intestinal perforation. PLoS One 2015;10:e0135737.
heart disease. J Pediatr 1988;113:1044–6. [45] Mohamed A, Shah PS. Transfusion associated necrotizing enterocolitis: a meta-
[19] Ostlie DJ, Spilde TL, St Peter SD, et al. Necrotizing enterocolitis in full-term infants. analysis of observational data. Pediatrics 2012;129:529–40.
J Pediatr Surg 2003;38:1039–42. [46] Blau J, Calo JM, Dozor D, Sutton M, Alpan G, La Gamma EF. Transfusion-related
[20] McElhinney DB, Hedrick HL, Bush DM, et al. Necrotizing enterocolitis in neonates acute gut injury: necrotizing enterocolitis in very low birth weight neonates after
with congenital heart disease: risk factors and outcomes. Pediatrics packed red blood cell transfusion. J Pediatr 2011;158:403–9.
2000;106:1080–7. [47] Kirpalani H, Zupancic JA. Do transfusions cause necrotizing enterocolitis? The
[21] Pickard SS, Feinstein JA, Popat RA, Huang L, Dutta S. Short- and long-term out- complementary role of randomized trials and observational studies. Semin Perinatol
comes of necrotizing enterocolitis in infants with congenital heart disease. 2012;36:269–76.
Pediatrics 2009;123:e901–6. [48] Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating
[22] Carlo WF, Kimball TR, Michelfelder EC, Border WL. Persistent diastolic flow re- quality of evidence and strength of recommendations. BMJ 2008;336:924–6.
versal in abdominal aortic Doppler-flow profiles is associated with an increased risk [49] Hay S, Zupancic JA, Flannery DD, Kirpalani H, Dukhovny D. Should we believe in
of necrotizing enterocolitis in term infants with congenital heart disease. Pediatrics transfusion-associated enterocolitis? Applying a GRADE to the literature. Semin
2007;119:330–5. Perinatol 2017;41:80–91.
[23] Powell GK. Enterocolitis in low-birth-weight infants associated with milk and soy [50] Patel RM, Knezevic A, Shenvi N, et al. Association of red blood cell transfusion,
protein intolerance. J Pediatr 1976;88:840–4. anemia, and necrotizing enterocolitis in very low-birth-weight infants. J Am Med
[24] Srinivasan P, Brandler M, D'Souza A, Millman P, Moreau H. Allergic enterocolitis Assoc 2016;315:889–97.
373

A critical analysis of risk factors for necrotizing enterocolitis T

∗
Allison Thomas Rose, Ravi Mangal Patel
Emory University School of Medicine and Children's Healthcare of Atlanta, Atlanta, GA, USA
Keywords: Necrotizing enterocolitis (NEC) is the most common serious gastrointestinal morbidity in preterm infants. A
Necrotizing enterocolitis number of risk factors for NEC have been reported in the literature. With the exception of decreasing gestational
Risk factor age, decreasing birth weight and formula feeding, there is disagreement on the importance of reported risk
Prematurity factors with uncertain causality. Causal risk factors may be observed at any time before the onset of NEC,
Epidemiology
including prior to an infant's birth. The purpose of this review is to examine the existing literature and sum-
Neonate
marize risk factors for NEC. This review may be helpful in understanding the epidemiology of NEC and inform
the measurement and assessment of risks factors for NEC in research studies and quality improvement projects.
1. Introduction (RCTs) and multicenter or regional/national observational studies

specifically assessing risk factors for NEC were reviewed, including
Necrotizing enterocolitis (NEC) is a serious and deadly gastro- reference lists. Risk factors are presented as they occur in sequence
intestinal disease with a multifactorial etiology that primarily occurs in during the fetal and neonatal life course: maternal and antenatal,
preterm infants. Data from the US show a decreasing incidence of NEC perinatal and delivery, early neonatal, and postnatal (including enteral
[1–3]. Among high-income countries, there is a large variation in the feeding). This review focuses on NEC as it develops in moderately or
incidence of NEC with differences ranging up to five-fold, from less than extremely preterm infants, and therefore does not discuss factors un-
2% to almost 9% for infants < 1500 g [4]. However, comparison of ique to more mature late-preterm and term infants. Given the distinc-
studies is limited by different populations and different definitions of tion between factors associated with NEC and factors that cause NEC,
NEC; this variation, however, suggests that modifiable risk factors exist we distinguished when data were from randomized trials in comparison
that could allow clinicians and researchers the opportunity to intervene to observational studies and when findings were consistent between the
to reduce the risk of NEC. This potential opportunity is reinforced by two study types.
large-scale, multifaceted quality improvement projects that have de-
monstrated decreases in NEC incidence associated with implementation 3. Maternal and antenatal
of targeted interventions [3]. However, clinicians and experts disagree
about the importance of different risk factors. In a group of 35 inter- Multiple studies have shown that NEC rates vary by region, net-
national NEC experts asked to review 64 identified risk factors, the only work, and country [1,2,4]; ethnic, racial, and genetic factors specific to
factors for which there was high agreement were gestational age, birth a given region are likely non-modifiable risk factors for certain infants.
weight and feeding (formula vs breastfeeding) [5]. The purpose of this In US studies, black race has consistently been associated with NEC
review is to evaluate the evidence on risk factors for NEC and to [6–8] as well as death from NEC (risk of death from NEC for black
summarize the consistency and direction of association of commonly infants compared to white infants in the USA, relative risk (RR): 3.1;
reported risk factors with NEC (Table 1). 95% confidence interval (CI): 3.0–3.4) [9]. Other maternal demo-
graphic factors including age, weight, multiparity, smoking, education,
2. Approach and employment status have not been consistently associated with NEC.
There are conflicting data on the role of hypertensive disorders of
To identify studies, we used the search term “necrotizing en- pregnancy and NEC risk. A meta-analyses including observational stu-
terocolitis” or “necrotising enterocolitis” in PubMed with the search dies between 2000 and 2016 showed no difference in NEC risk between
limited to systematic reviews. Additional randomized controlled trials all mothers with pregnancy-induced hypertensive disorders and those
∗
Corresponding author. Emory University School of Medicine and Children's Healthcare of Atlanta, 2015 Uppergate Drive, NE, Atlanta, GA, 30322, USA. Tel.: +1
404 727 5905; fax: +1 404 727 3236.
E-mail address: rmpatel@emory.edu (R.M. Patel).
A.T. Rose, R.M. Patel Seminars in Fetal and Neonatal Medicine 23 (2018) 374–379
Table 1 0.93–4.79; three studies, 395 infants) [11].

Summary of risk factors for necrotizing enterocolitis (NEC). Antenatal steroid (ANS) administration reduces neonatal mortality;
Factor Association Observational Randomized a meta-analysis of 10 randomized controlled trials showed that ANS
with NEC study controlled trial decreased the relative risk of NEC by 50% (RR: 0.50; 95% CI:
0.32–0.78; 10 studies, 4702 infants) [12]. However, a meta-analysis of
Maternal and antenatal
observational studies showed no benefit of ANS on NEC at very low
Black race ↑ X
Maternal ∼ X X
gestational ages (OR: 1.01; 95% CI: 0.84–1.22; seven studies, 8737
hypertension infants < 25 weeks gestation), potentially due to effects on the com-
Pre-eclampsia ↑ X peting cause of death (i.e. more survivors increase the number of in-
Antenatal steroids ↓ X fants at risk for NEC) [13].
Maternal ↓ X
Some therapies given to mothers have been shown to decrease the
progesterone
Tocolytics ~ X X risk of NEC. Progesterone given to women at risk for preterm birth
Magnesium sulfate - X decreased the relative risk of NEC by 70% (RR: 0.30; 95% CI:
Maternal infection ↑ X 0.10–0.89; three studies, 1170 infants) [14]. A review of RCT data
Prolonged rupture of ↑ X
examining the use of calcium channel blockers for tocolysis compared
membranes
Maternal antibiotics ↓/~ X X
to beta-mimetics showed a reduced risk of NEC (RR: 0.21; 95% CI:
Placental abruption ↑ X 0.05–0.96; five studies, 490 infants), which may be due to increases in
Intrauterine growth ↑ X gestational age at birth from pregnancy prolongation [15]. However,
restriction meta-analysis of observational studies suggests that the use of in-
Infant baseline and delivery
domethacin for tocolysis may increase the risk of NEC (RR: 1.36; 95%
Cesarean delivery ∼ X X
Delayed cord - X CI: 1.08–1.71; 18 studies, 4273 infants) [16].
clamping Magnesium sulfate has several indications for use in obstetrics, in-
Younger gestational ↑ X cluding pre-eclampsia or eclampsia, fetal neuroprotection in threatened
age preterm delivery (< 32 weeks gestation), and short-term prolongation
Lower birth weight ↑ X
Small for gestational ↑ X
of pregnancy to allow for administration of ANS. Indirect evidence from
age RCTs of magnesium sulfate to reduce cerebral palsy in premature in-
Male sex ∼ X fants (24–31 weeks gestation) shows no significant increased risk of
Postnatal NEC for those receiving magnesium prior to delivery (RR: 1.27; 95% CI:
Surfactant ↑/~ X X
0.97–1.66; 2415 infants) [17].
Lower O2 target ↑ X
Umbilical catheters ∼ X X Maternal chorioamnionitis, both clinically and histologically de-
Patent ductus ↑/~ X X fined, has been examined with varying results on its association with
arteriosus NEC. A 2013 meta-analysis found histological chorioamnionitis with
Sepsis ↑ X fetal involvement (i.e. maternal and concomitant fetal inflammation)
Empiric antibiotics ↑/~ X X
Severity of anemia ↑/~ X X
was associated with a three-fold increased risk for NEC (OR: 3.29; 95%
Red blood cell ∼ X X CI: 1.87–5.78) [18]. Prolonged rupture of membranes, a common
transfusion antecedent to chorioamnionitis, has been associated with NEC across
Erythropoietin ↓ X several studies spanning several decades [19–21]. Whereas maternal
Fluid restriction ↓ X X
infection may increase NEC risk for the infant, maternal antibiotic
Oral lactoferrin ↓ X
Arginine ↓ X treatment does not appear to increase risk and may, in fact, decrease it.
supplementation Data from a prospective, observational microbiome study found a re-
Glycerin enemas/ ∼ X duced risk of NEC if maternal antibiotics were given (OR: 0.28; 95% CI:
suppositories 0.14–0.56; 580 infants) [22]. However, meta-analysis of RCT data of
Feeding
Breast milk ↓ X X
maternal antibiotics in the setting of premature rupture of membranes
Preterm formula ↑ X X showed no effect of antibiotic administration on NEC (RR: 1.09; 95%
Donor human milk ↓ X X CI: 0.65–1.83; 11 studies, 6229 infants), except with the use of amox-
Delayed feeding - X icillin–clavulanate (RR: 4.72; 95% CI: 1.57–14.23; two studies, 1880
Slow feeding - X
infants) [23]. Differences between these studies may be due to residual
Trophic feeding - X
Feeding protocol ↓ X confounding, differences in the populations, including presence or ab-
Probiotics ↓ X X sence of premature rupture of membranes (PROM), and ascertainment
Acid suppression ↑ X of chorioamnionitis.
Assessment of risk factors based on studies identified in search strategy. Please

4. Perinatal and delivery characteristics
see text for additional details.
↑: factor associated with increased risk of NEC; ↓: decreased risk of NEC; -: no
evidence of effect on NEC; ∼: indeterminate effect on NEC. If multiple symbols
4.1. Fetal ischemia and growth
present, indicates conflicting data from observational studies (first symbol) and
randomized controlled trials (second symbol). Compromised fetal blood flow before or at the time of delivery may
result in fetal ischemia that could contribute to NEC. An Australian
without (OR: 1.89; 95% CI: 0.82–4.37; four studies, 28,724 infants) cohort of infants 24–31 weeks gestation had a two-fold increased risk of
[10]. However, in the study's subgroup analysis of mothers with pre- NEC in the setting of placental abruption (aOR: 2.09; 95% CI:
eclampsia, there was a 2.5-fold increased risk for NEC based on data 1.30–3.35; 4649 infants) [24]. An observational study found a similar
from three studies (878 infants). Analysis of three RCTs of interven- association with vaginal bleeding after 12 weeks of gestation and NEC
tional management (e.g. delivery), compared to expectant management [6].
for pre-eclampsia between 24 and 34 weeks, showed a non-significant Infants with intrauterine growth restriction (IUGR), with or without
trend toward increased NEC among infants in the intervention group, abnormal Doppler studies, and small for gestational age (SGA) infants
likely due to delivery at an earlier gestational age (RR: 2.10; 95% CI: are at increased risk for NEC [24–26]. A 2005 meta-analysis of ob-
servational studies of infants with absent end-diastolic flow (AEDF)
375
versus controls showed increased odds of NEC in the AEDF group (OR: (RR: 0.20; 95% CI: 0.08–0.51) [40]. Animal-derived surfactant, how-
2.13; 95% CI: 1.49–3.03; 14 studies, 1837 infants) [27]. ever, has been associated with an increased risk of NEC (RR: 1.38; 95%
CI: 1.08–1.76; eight studies 3462 infants) [41]. In addition, lower
4.2. Delivery oxygen saturation targets increase the risk of NEC in extremely preterm
infants; the Neonatal Oxygen Prospective Meta-analysis Collaborative
Approximately 60% of extremely preterm infants are delivered by showed that NEC risk increased for infants < 28 weeks with a lower
cesarean section [1]. Studies have reported conflicting data regarding oxygen saturation goal of 85–89% compared to those in the 91–95%
the relationship of cesarean delivery to risk of NEC with studies target group (RR: 1.24; 95% CI: 1.05–1.47; five studies, 4929 infants)
showing cesarean to be protective [8,20], not associated [28], or as- [42].
sociated [19] with increased NEC risk. Secondary analysis of RCT data The presence of an indwelling umbilical arterial catheter has been
showed no association with cesarean delivery and NEC (OR: 0.73; 95% an exclusion criterion for infants in feeding advancement studies, sug-
CI: 0.46–1.16; 2012 infants) [29]. Emerging literature suggests the gesting concern that these catheters may put infants at risk for NEC
potential for the fetal gut microbiome to develop before delivery [30]. [43]. However, there is inconsistent evidence of the risk associated with
Therefore, the effect of mode of delivery on NEC and related changes to umbilical catheters [8,36] and the location of umbilical artery cathe-
the gut microbiome is likely complex and multifactorial. ters, with a meta-analysis of RCTs showing no effect of umbilical ar-
The evidence for a reduced risk of NEC in premature infants who terial catheters on the risk of NEC (RR: 1.34; 95% CI: 0.79–2.25; five
receive delayed cord clamping has recently changed. A meta-analysis studies, 1569 infants) [44].
that includes the Australian Placental Transfusion Study [31] found no
effect of delayed cord clamping on NEC in infants born < 37 weeks
gestation or in a subgroup of those born ≤28 weeks (OR: 0.88; 95% CI: 6. Postnatal factors
0.65–1.18; 12 studies, 2397 infants; and OR: 0.87; 95% CI: 0.61–1.24;
four studies, 977 infants, respectively) [32]. These data contrast with a 6.1. Complications associated with prematurity
prior meta-analysis of 241 infants showing a 40% reduction in NEC for
infants receiving delayed cord clamping [33]. A patent ductus arteriosus (PDA) may be present in up to 65% of
infants born < 29 weeks gestation [45] and treatment for a PDA may
5. Baseline or early neonatal occur prior to the development of NEC. In observational studies, it can
be difficult to determine whether the presence of a PDA or the treat-
5.1. Baseline ment of the PDA modifies an infant's risk of NEC. In a cohort of in-
fants < 34 weeks in Israel, those with a PDA without indomethacin
Decreasing gestational age and birth weight are consistently re- treatment and those with a PDA with indomethacin treatment had an
ported in the literature as clear risk factors for NEC [1]. In a national increased risk for NEC compared to infants without a PDA (OR: 1.85;
study of infants in the UK, the incidence of NEC was 11% in those born 95% CI: 1.24–2.69; and OR: 1.53; 95% CI: 1.15–2.02, respectively;
at 24 weeks gestation and decreased to 0.5% for infants born at 31 6044 infants) [46]. In a recent network meta-analysis of randomized
weeks gestation [34]. This study also demonstrated an increased risk trials, there was no increased risk of NEC with no treatment (or pla-
for NEC in the setting of decreasing birth weight z-score (adjusted (a) cebo) for PDA; the analysis found that some pharmacological regimens
OR: 1.29; 95% CI: 1.17–1.43 for each unit decrease in birth weight z- performed worse than placebo/no treatment, although the differences
score). A large national study in the USA had similar findings and de- did not reach statistical significance [47].
monstrated an increased risk of death from NEC in the setting of de- Sepsis places an infant at risk for NEC [21,37,38]. However, empiric
creasing gestational age and decreasing birth weight [7]. antibiotics and the duration of antibiotic exposure has unclear effects
There have not been consistent findings for the association between on NEC risk. Several studies have found that the duration of empiric
infant sex and NEC. A study of incidence of NEC in Sweden found male antibiotics in extremely low birth weight infants with sterile cultures is
sex to be associated with NEC; however, a case–control, study during associated with an increased risk of NEC [48]. A recent meta-analysis of
the same time period, did not associate male sex with an increased risk observational studies found prolonged antibiotic exposure was asso-
of NEC [19,35]. Because of the established association of sex and ciated with NEC in 5 studies (5003 infants) but among RCTs found no
mortality in extremely preterm infants, consideration of the association statistically significant effect on NEC when looking at whether or not an
between sex and NEC would require an understanding of changes in the infant received prophylactic antibiotics (OR: 1.25; 95% CI: 0.12–12.50;
competing causes of death within a given cohort and also potential two RCT, 288 infants) or broad-versus narrow-spectrum antibiotics
interaction with other factors such as race or birth weight. (OR: 0.30; 95% CI: 0.04–2.31; three RCT, 586 infants) [49]; however,
Multiple measures of poor neonatal transition have been associated the trend was toward increased risk of NEC. As all of the data regarding
with NEC, including need for resuscitation in the delivery room, low duration of antibiotics is from observational studies, the estimates of
Apgar score, hypotension, low umbilical cord pH or base deficit, and association with NEC from these studies could be biased by residual
early mechanical ventilation [8,36–38]. These factors are primarily confounding.
reported in observational studies and have not been consistently asso- Anemia, with the potential for poor intestinal oxygen delivery, has
ciated with NEC, raising questions regarding whether they are a cause been studied as a risk factor for NEC. A prospective observational study
of NEC or only associated with NEC. Regardless, the associations sug- found that anemia (≤8 g/dL) was associated with an increased risk for
gest the potential importance of early physiology stability that could NEC (hazard ratio (HR): 5.99; 95% CI: 2.00–18.0) but not for red cell
contribute to the subsequent risk of NEC. transfusion (HR: 0.44; 95% CI: 0.17–1.12) [50]. Data from RCTs
showed no difference in the risk of NEC for infants transfused using a
5.2. Early neonatal lower versus higher hemoglobin threshold (RR: 1.62; 95% CI:
0.83–3.13; three studies, 590 infants) [51]. The risk of red cell trans-
Whereas some large observational studies report on the positive fusion is discussed below and additional data from the Transfusion of
association between surfactant and NEC [8,24], a meta-analysis of Prematures Trial (clinicaltrials.gov NCT01702805) may provide addi-
randomized trials of prophylactic and selective surfactant use found no tional insight into the relationship between red cell transfusion, an-
association with NEC (RR: 0.90; 95% CI: 0.73–1.10; eight studies, 4237 emia, and NEC.
infants) [39]. In addition, the administration of multiple doses of sur-
factant, compared to a single dose, resulted in a decreased risk of NEC
376
6.2. Care interventions recent RCT of supplemental fortified donor human milk versus preterm
formula found a decreased risk of NEC (all stages and ≥II) in the donor
Whether red cell transfusion is a causal risk factor for NEC is not milk group (risk difference: −7.1; 95% CI: −12.5 to −1.8 and risk
currently known. A recent meta-analysis shows no increased risk of NEC difference: −4.9; 95% CI: −9.0 to −0.9, respectively, 361 infants)
following red cell transfusion (OR: 1.13; 95% CI: 0.99–1.29; 13 ob- [66]. Bovine product exposure among human-milk-fed infants may be
servational trials, 11,602 infants) [52] and another suggests that associated with a small increased risk of NEC [34]. At this time, there
transfusion may be protective (OR: 0.55; 95% CI: 0.31–0.98; 10 ob- are no data on oropharyngeal colostrum administration and the risk of
servational trials, 24,623 infants) [53]. Observational data suggest that NEC.
feeding during transfusion is a risk factor for NEC within 48 h (RR NEC, Clinical instability and concerns for immature gut motility often
NPO vs feeding: 0.47; 95% CI: 0.28–0.80; seven studies, 7492 infants) lead to a delay in enteral feeding. Meta-analysis has shown that delayed
[54]. feeding in infants < 1500 g or < 32 weeks does not alter NEC risk (RR:
Erythropoietin may be protective for endothelial cell barriers and, 0.93; 95% CI: 0.64–1.34; eight studies, 1092 infants), including among
thus, may mitigate the development of NEC. A meta-analysis of RCTs of IUGR infants with abnormal umbilical Doppler flow velocities (RR:
early (< 8 days after birth) erythropoiesis-stimulating agents used to 0.87; 95% CI: 0.54–1.41) [67]. The studies included in the meta-ana-
modify transfusion exposure in very-low-birth-weight infants demon- lysis had few extremely low birth weight or extremely preterm infants
strated a decrease in NEC risk in those infants receiving the intervention and planned subgroup analysis of this cohort was not performed.
(RR: 0.69; 95% CI: 0.52–0.91; 15 studies, 2639 infants) [55]. Therefore, the best feeding strategy for the most immature infants re-
Total fluid goals for premature infants are influenced by multiple mains unclear. Studies have incompletely examined trophic feeding
factors but a restricted approach may decrease NEC risk. In a meta- versus enteral fasting in extremely low birth weight infants, but meta-
analysis of RCTs, restricted versus liberal fluid intake reduced the risk analysis of infants < 1500 g suggests that there is no increased risk of
of NEC in infants < 2000 g (RR: 0.43; 95% CI: 0.21–0.87; four studies, NEC with trophic feeds (RR: 1.07; 95% CI: 0.67–1.70) [68]. Once feeds
526 infants) [56]. An multicenter observational study found that cen- are initiated, there is no difference in NEC risk for infants advanced at a
ters more likely to restrict fluids also had the lowest NEC incidence slow rate (10–20 mL/kg/day) versus a faster rate (30–40 mL/kg/day)
[20]. [43]. This finding is consistent in multiple subgroup analysis including
Caffeine and postnatal corticosteroids are widely used treatments groups of extremely low birth weight or extremely preterm infants, SGA
for respiratory morbidity. A large randomized trial found that caffeine or IUGR infants, infants with absent or reversed end-diastolic flow,
is not a risk factor for NEC [57]; current studies examining early versus infants fed mostly formula, and infants fed partially human milk [43].
later caffeine use also report no difference in NEC [58]. A meta-analysis Standardized feeding regimens have been associated with a reduc-
of timing of postnatal corticosteroids, dosing regimens, and methods tion in risk of NEC. A 2017 systematic review of observational studies
suggests that postnatal steroid use is not a risk factor for NEC [59]. found an 80% reduction in relative risk of NEC when comparing stan-
dardized feedings with non-standardized feedings (RR: 0.22; 95% CI:
6.3. NEC-specific treatments 0.13–0.36; 15 studies, 18,160 infants). This benefit persisted in analysis
limited to recent years (RR: 0.26; 95% CI: 0.19–0.35; nine studies, 9456
Several RCTs have reported on the effects of oral lactoferrin, a infants) [69]. Feeding regimens have varied in approaches, although all
naturally occurring glycoprotein with antimicrobial and im- recommend breastfeeding. Approaches to feeding during transfusions
munomodulatory properties. A meta-analysis of four RCTs found a have been discussed previously.
protective effect of oral lactoferrin over placebo or no intervention (RR: There have been numerous studies of the role of probiotics in de-
0.40; 95% CI: 0.18–0.86; four studies, 750 infants) [60]. Similar results creasing the risk of NEC. Whereas questions remain about optimal
were seen when oral lactoferrin was combined with probiotics [60]. strain, dose, duration, and product quality, meta-analyses have shown
Specific amino acid supplementation has been proposed to alter consistent results on the effect of probiotics in decreasing the risk of
NEC incidence. Arginine may alter nitric oxide metabolism, which is NEC [70]. Further knowledge and understanding of the premature in-
thought to play a role in gut perfusion and motility. Meta-analysis of fant's microbiome may help target probiotic-like therapies. A meta-
small trials demonstrates that oral or parenteral supplementation with analysis of stool microbiome profiles in premature infants revealed a
arginine decreases the risk for any stage of NEC (RR: 0.38; 95% CI: relative increase in Proteobacteria and a relative decrease in Firmicutes
0.23–0.64; three studies, 285 infants) and death due to NEC (RR: 0.18; and Bacteroidetes in NEC cases compared to controls [71].
94% CI: 0.03–1.00) [61]. Whereas the administration of probiotics decreases the risk of NEC,
Rectal administration of osmotic agents is frequently used to pre- enteral administration of medications that promote acid suppression
vent or treat feeding intolerance, lack of stooling or abdominal dis- has the opposite effect. Pooled analysis from two observational trials,
tention. Their prophylactic use does not appear to impact NEC risk. In a one case–control and one prospective, shows a significantly increased
meta-analysis of six studies using different forms of meconium eva- risk of NEC with acid-suppression medications (OR: 1.78; 95% CI:
cuation strategies, there was no effect on NEC (RR: 1.71; 95% CI: 1.4–2.27; two studies, 11,346 infants) [72].
0.63–4.65; six studies, 442 infants) but there was a trend towards in-
creasing risk with oral gastrographin administration (RR: 2.61; 95% CI: 8. Risk factor scoring tools
0.88–7.74; one study, 96 infants) [62]. A review of only glycerin
laxative enemas or suppositories showed no effect on NEC (RR: 2.75; Necrotizing enterocolitis risk factor scoring tools have been devel-
95% CI: 0.58–13.1; two studies 96 infants), although the point estimate oped to identify at-risk infants. GutCheckNEC was developed with evi-
suggested a greater probability of harm than benefit in the risk of NEC dence review, international expert consensus, and statistical modeling;
with use of glycerin [63]. it has been validated with existing multi-institutional data [5,38]. Unit
NEC rate is the most heavily weighted variable when determining risk
7. Feeding status (accounting for nine to 23 of 32 necessary points used to consider
risk of NEC). eNEC uses similar inputs as GutCheck, was developed in a
Breastfeeding has been consistently shown to decrease the risk of single institution setting, and was designed to be completed on ad-
NEC [64]. When maternal breast milk is unavailable, preterm formula, mission and on a weekly basis. It has been studied for feasibility but not
compared to donor breast milk, increases the risk of NEC (RR: 2.61; yet for outcomes [73]. NeoNEEDS differs from both the previous tools
95% CI: 1.27–5.35; five studies, 802 infants) [65], but this meta-ana- as it determines a risk status based on the infant's current clinical
lysis only contained one trial (173 infants) using fortified donor milk. A condition in five domains (behavior, cardiovascular, respiratory,
377
abdominal and feeding) combined with gestational age or birth weight Syst Rev 2013;7:CD003106.
and exposure to formula [74]. It has significantly fewer inputs than the [12] Roberts D, Brown J, Medley N, Sr D. Antenatal corticosteroids for accelerating fetal
lung maturation for women at risk of preterm birth (Review). Cochrane Database
other models and was designed as part of a quality improvement project Syst Rev 2017;1:1–273.
to be completed daily during care team rounds. An at-risk score triggers [13] Deshmukh M, Patole S. Antenatal corticosteroids in impending preterm deliveries
the clinician to consider ordering an abdominal X-ray. NEC severity before 25 weeks' gestation. Arch Dis Child Fetal Neonatal Ed 2018;103:F173–6.
[14] Dodd JM, Jones L, Flenady V, Cincotta R, Crowther CA. Prenatal administration of
rates decreased during the study period at a single site using this tool progesterone for preventing preterm birth in women considered to be at risk of
[74]. preterm birth. Cochrane Database Syst Rev 2013;7:CD004947.
[15] Flenady V, Wojcieszek AM, Papatsonis DNM, et al. Calcium channel blockers for
inhibiting preterm labour and birth. Cochrane Database Syst Rev
9. Summary 2014;6:CD002255.
[16] Hammers AL, Sanchez-Ramos L, Kaunitz AM. Antenatal exposure to indomethacin
Various factors are associated with the risk of NEC reported in the increases the risk of severe intraventricular hemorrhage, necrotizing enterocolitis,
and periventricular leukomalacia: a systematic review with metaanalysis. Am J
literature. The interaction between these factors, the role of center, and
Obstet Gynecol 2015;212:505. e1–13.
the effect of residual confounding limit the interpretation of causality [17] Rouse DJ, Hirtz DG, Thom E, et al. A randomized, controlled trial of magnesium
and the role of these factors for any given infant's risk of NEC. sulfate for the prevention of cerebral palsy. N Engl J Med 2008;359:895–905.
Nevertheless, this review found many potentially modifiable risk factors [18] Been JV, Lievense S, Zimmermann LJI, Kramer BW, Wolfs TGAM. Chorioamnionitis
as a risk factor for necrotizing enterocolitis: a systematic review and meta-analysis.
that could be targets of research and quality improvement efforts. In J Pediatr 2013;162. 236–242.e2.
addition, the summary of risk factors may be useful to ensure appro- [19] Ahle M, Drott P, Elfvin A, Andersson RE. Maternal, fetal and perinatal factors as-
priate measurement and assessment of factors in clinical studies. sociated with necrotizing enterocolitis in Sweden. A national case–control study.
PLoS One 2018;13:e0194352.
[20] Uauy R, Fanaroff A, Korones S, Phillips E, Phillips J, Wright L. Necrotizing en-
Practice points terocolitis in very low birth weight infants: biodemographic and clinical correlates.
J Pediatr 1991;119:630–8.
• Factors that increase or decrease the risk of NEC may be present [21] Drenckpohl D, Knaub L, Schneider C, McConnell C, Wang H, Macwan K. Risk factors
that may predispose premature infants to increased incidence of necrotizing en-
prior to birth through the postnatal course. terocolitis. Infant Child Adolesc Nutr 2010;2:37–44.
• Studies of NEC should consider measurement of important risk [22] Reed BD, Schibler KR, Deshmukh H, Ambalavanan N, Morrow AL. The impact of
maternal antibiotics on neonatal disease. J Pediatr 2018;197:97–103.e3.
factors to ensure appropriate balance of covariates in randomized [23] Kenyon S, Boulvain M, Neilson JP. Antibiotics for preterm rupture of membranes.
trials or assessment of confounding in observational studies. Cochrane Database Syst Rev 2013;12. CD001058.
• Additional studies are needed to determine the interaction between [24] Luig M, Lui K, NSW & ACT NICUS Group. Epidemiology of necrotizing enterocolitis
– Part II: risks and susceptibility of premature infants during the surfactant era: a
risk factors and the influence of center.
•
regional study. J Paediatr Child Health 2005;41:174–9.
The causal structure of a variable's relationship to NEC is important [25] Garite TJ, Clark R, Thorp JA. Intrauterine growth restriction increases morbidity
to consider when determining whether a specific risk factor causes and mortality among premature neonates. Am J Obstet Gynecol 2004;191:481–7.
NEC. [26] Bernstein IM, Horbar JD, Badger GJ, Ohlsson A, Golan A. Morbidity and mortality
among very-low-birth-weight neonates with intrauterine growth restriction. The
Vermont Oxford Network. Am J Obstet Gynecol 2000;182(1 Pt 1):198–206.
Conflict of interest statement [27] Dorling J, Kempley S, Leaf A. Feeding growth restricted preterm infants with ab-
normal antenatal Doppler results. Arch Dis Child Fetal Neonatal Ed
2005;90:F359–63.
None declared. [28] Battersby C, Longford N, Costeloe K, Modi N. Development of a gestational age-
specific case definition for neonatal necrotizing enterocolitis. JAMA Pediatr
Funding source 2017;171:256.
[29] Son M, Grobman WA, Miller ES. Is mode of delivery associated with the risk of
necrotizing enterocolitis? Am J Obstet Gynecol 2016;215:389. e1–4.
Dr Patel received support from the National Institutes of Health [30] Aagaard K, Ma J, Antony KM, Ganu R, Petrosino J, Versalovic J. The placenta
under award number K23 HL128942. harbors a unique microbiome. Sci Transl Med 2014;6. 237ra65.
[31] Tarnow-Mordi W, Morris J, Kirby A, et al. Delayed versus immediate cord clamping
in preterm infants. N Engl J Med 2017;377:2445–55.
References [32] Fogarty M, Osborn DA, Askie L, et al. Delayed vs early umbilical cord clamping for
preterm infants: a systematic review and meta-analysis. Am J Obstet Gynecol
2018;218:1–18.
[1] Stoll BJ, Hansen NI, Bell EF, et al. Trends in care practices, morbidity, and mortality
[33] Rabe H, Diaz-Rossello JL, Duley L, Dowswell T. Effect of timing of umbilical cord
of extremely preterm neonates, 1993–2012. J Am Med Assoc 2015;314:1039.
clamping and other strategies to influence placental transfusion at preterm birth on
[2] Horbar JD, Edwards EM, Greenberg LT, et al. Variation in performance of neonatal
maternal and infant outcomes. Cochrane Database Syst Rev 2012;8:CD003248.
intensive care units in the United States. JAMA Pediatr 2017;171:e164396.
[34] Battersby C, Longford N, Mandalia S, Costeloe K, Modi N. Incidence and enteral
[3] Ellsbury DL, Clark RH, Ursprung R, Handler DL, Dodd ED, Spitzer AR.
feed antecedents of severe neonatal necrotising enterocolitis across neonatal net-
A multifaceted approach to improving outcomes in the NICU: the Pediatrix 100 000
works in England, 2012–13: a whole-population surveillance study. Lancet
Babies Campaign. Pediatrics 2016;137. e20150389–e20150389.
Gastroenterol Hepatol 2017;2:43–51.
[4] Battersby C, Santhalingam T, Costeloe K, Modi N. Incidence of neonatal necrotising
[35] Ahle M, Drott P, Andersson RE. Epidemiology and trends of necrotizing en-
enterocolitis in high-income countries: a systematic review. Arch Dis Child Fetal
terocolitis in Sweden: 1987–2009. Pediatrics 2013;132:e443–51.
Neonatal Ed 2018;103:F182–9.
[36] Lee J-Y, Park K-H, Kim A, Yang H-R, Jung E-Y, Cho S-H. Maternal and placental risk
[5] Gephart SM, Effken JA, McGrath JM, Reed PG. Expert consensus building using e-
factors for developing necrotizing enterocolitis in very preterm infants. Pediatr
Delphi for necrotizing enterocolitis risk assessment. J Obstet Gynecol Neonatal Nurs
Neonatol 2017;58:57–62.
2013;42:332–47.
[37] Gagliardi L, Bellù R, Cardilli V, et al. Necrotising enterocolitis in very low birth
[6] Singh R, Shah B, Allred EN, Grzybowski M, Martin CR, Leviton A. The antecedents
weight infants in Italy: incidence and non-nutritional risk factors. J Pediatr
and correlates of necrotizing enterocolitis and spontaneous intestinal perforation
Gastroenterol Nutr 2008;47:206–10.
among infants born before the 28th week of gestation. J Neonatal Perinat Med
[38] Gephart SM, Spitzer AR, Effken JA, Dodd E, Halpern M, McGrath JM.
2016;9:159–70.
Discrimination of GutCheck(NEC): a clinical risk index for necrotizing enterocolitis.
[7] Clark RH, Gordon P, Walker WM, Laughon M, Smith PB, Spitzer AR. Characteristics
J Perinatol 2014;34:468–75.
of patients who die of necrotizing enterocolitis. J Perinatol 2012;32:199–204.
[39] Rojas-Reyes MX, Morley CJ, Soll R. Prophylactic versus selective use of surfactant in
[8] Guthrie SO, Gordon PV, Thomas V, Thorp JA, Peabody J, Clark RH. Necrotizing
preventing morbidity and mortality in preterm infants. Cochrane Database Syst Rev
enterocolitis among neonates in the United States. J Perinatol 2003;23:278–85.
2012;3:CD000510.
[9] Seeman SM, Mehal JM, Haberling DL, Holman RC, Stoll BJ. Infant and maternal risk
[40] Corbet A, Gerdes J, Long W, et al. Double-blind, randomized trial of one versus
factors related to necrotising enterocolitis-associated infant death in the United
three prophylactic doses of synthetic surfactant in 826 neonates weighing 700 to
States. Acta Paediatr 2016;105:e240–6.
1100 grams: effects on mortality rate. American Exosurf Neonatal Study Groups I
[10] Razak A, Florendo-Chin A, Banfield L, et al. Pregnancy-induced hypertension and
and IIa. J Pediatr 1995;126:969–78.
neonatal outcomes: a systematic review and meta-analysis. J Perinatol
[41] Ardell S, Pfister RH, Soll R. Animal derived surfactant extract versus protein free
2018;38:46–53.
synthetic surfactant for the prevention and treatment of respiratory distress syn-
[11] Churchill D, Duley L, Thornton JG, Jones L. Interventionist versus expectant care
drome. Cochrane Database Syst Rev 2015;8:CD000144.
for severe pre-eclampsia between 24 and 34 weeks' gestation. Cochrane Database
[42] Askie LM, Darlow BA, Davis PG, et al. Effects of targeting lower versus higher
378
arterial oxygen saturations on death or disability in preterm infants. Cochrane [59] Onland W, De Jaegere AP, Offringa M, van Kaam A. Systemic corticosteroid regi-
Database Syst Rev 2017;4:CD011190. mens for prevention of bronchopulmonary dysplasia in preterm infants. Cochrane
[43] Oddie S, Young L, McGuire W. Slow advancement of enteral feed volumes to pre- Database Syst Rev 2017;1:CD010941.
vent necrotising enterocolitis in very low birth weight infants. Cochrane Database [60] Pammi M, Suresh G. Enteral lactoferrin supplementation for prevention of sepsis
Syst Rev 2017;10(8):CD001241. and necrotizing enterocolitis in preterm infants. Cochrane Database Syst Rev
[44] Barrington KJ. Umbilical artery catheters in the newborn: effects of position of the 2017;6:CD007137.
catheter tip. Cochrane Database Syst Rev 2000;2:CD000505. [61] Shah PS, Shah VS, Kelly LE. Arginine supplementation for prevention of necrotising
[45] Benitz WE. Patent ductus arteriosus in preterm infants. Pediatrics enterocolitis in preterm infants. Cochrane Database Syst Rev 2017;4:CD004339.
2016;137:e20153730. [62] Deshmukh M, Balasubramanian H, Patole S. Meconium evacuation for facilitating
[46] Dollberg S, Lusky A, Reichman B. Patent ductus arteriosus, indomethacin and ne- feed tolerance in preterm neonates: a systematic review and meta-analysis.
crotizing enterocolitis in very low birth weight infants: a population-based study. Neonatology 2016;110:55–65.
J Pediatr Gastroenterol Nutr 2005;40:184–8. [63] Anabrees J, Shah VS, AlOsaimi A, AlFaleh K. Glycerin laxatives for prevention or
[47] Mitra S, Florez ID, Tamayo ME, et al. Association of placebo, indomethacin, ibu- treatment of feeding intolerance in very low birth weight infants. Cochrane
profen, and acetaminophen with closure of hemodynamically significant patent Database Syst Rev 2015;9:CD010464.
ductus arteriosus in preterm infants: a systematic review and meta-analysis. J Am [64] Lucas A, Cole TJ. Breast milk and neonatal necrotising enterocolitis. Lancet
Med Assoc 2018;319:1221–38. 336(8730):1519–1523.
[48] Cotten CM, Taylor S, Stoll B, et al. Prolonged duration of initial empirical antibiotic [65] Quigley M, McGuire W. Formula versus donor breast milk for feeding preterm or
treatment is associated with increased rates of necrotizing enterocolitis and death low birth weight infants. Cochrane Database Syst Rev 2014;4. CD002971.
for extremely low birth weight infants. Pediatrics 2009;123:58–66. [66] O'Connor DL, Gibbins S, Kiss A, et al. Effect of supplemental donor human milk
[49] Esaiassen E, Fjalstad JW, Juvet LK, van den Anker JN, Klingenberg C. Antibiotic compared with preterm formula on neurodevelopment of very low-birth-weight
exposure in neonates and early adverse outcomes: a systematic review and meta- infants at 18 months: a randomized clinical trial. J Am Med Assoc
analysis. J Antimicrob Chemother 2017;72:1858–70. 2016;316:1897–905.
[50] Patel RM, Knezevic A, Shenvi N, et al. Association of red blood cell transfusion, [67] Morgan J, Young L, McGuire W. Delayed introduction of progressive enteral feeds
anemia, and necrotizing enterocolitis in very low-birth-weight infants. J Am Med to prevent necrotising enterocolitis in very low birth weight infants. Cochrane
Assoc 2016;315:889–97. Database Syst Rev 2014;12:CD001970.
[51] Whyte R, Kirpalani H. Low versus high haemoglobin concentration threshold for [68] Morgan J, Bombell S, McGuire W. Early trophic feeding versus enteral fasting for
blood transfusion for preventing morbidity and mortality in very low birth weight very preterm or very low birth weight infants. Cochrane Database Syst Rev
infants. Cochrane Database Syst Rev 2011;11. CD000512. 2013;3:CD000504.
[52] Hay S, Zupancic JAF, Flannery DD, Kirpalani H, Dukhovny D. Should we believe in [69] Jasani B, Patole S. Standardized feeding regimen for reducing necrotizing en-
transfusion-associated enterocolitis? Applying a GRADE to the literature. Semin terocolitis in preterm infants: an updated systematic review. J Perinatol
Perinatol 2017;41:80–91. 2017;37:827–33.
[53] Rai SE, Sidhu AK, Krishnan RJ. Transfusion-associated necrotizing enterocolitis re- [70] Patel RM, Underwood M. Probiotics and necrotizing enterocolitis. Semin Pediatr
evaluated: a systematic review and meta-analysis. J Perinat Med 2017 Oct 25. Surg 2018;27:39–46.
https://doi.org/10.1515/jpm-2017-0048. [Epub ahead of print]. [71] Pammi M, Cope J, Tarr PI, et al. Intestinal dysbiosis in preterm infants preceding
[54] Jasani B, Rao S, Patole S. Withholding feeds and transfusion-associated necrotizing necrotizing enterocolitis: a systematic review and meta-analysis. Microbiome
enterocolitis in preterm infants: a systematic review. Adv Nutr 2017;8:764–9. 2017;5:31.
[55] Ohlsson A, Aher SM. Early erythropoiesis-stimulating agents in preterm or low birth [72] More K, Athalye-Jape G, Rao S, Patole S. Association of inhibitors of gastric acid
weight infants. Cochrane Database Syst Rev 2017;11:CD004863. secretion and higher incidence of necrotizing enterocolitis in preterm very low-
[56] Bell EF, Acarregui MJ. Restricted versus liberal water intake for preventing mor- birth-weight infants. Am J Perinatol 2013;30:849–56.
bidity and mortality in preterm infants. Cochrane Database Syst Rev [73] Naberhuis J, Wetzel C, Tappenden KA. A novel neonatal feeding intolerance and
2014;12:CD000503. necrotizing enterocolitis risk-scoring tool is easy to use and valued by nursing staff.
[57] Schmidt B, Roberts R, Davis P. Caffeine therapy for apnea of prematurity. N Engl J Adv Neonatal Care 2016;16:239–44.
Med 2006;354:2112–21. [74] Fox JR, Thacker LR, Hendricks-Muñoz KD. Early detection tool of intestinal dys-
[58] Park HW, Lim G, Chung S-H, Chung S, Kim KS, Kim S-N. Early caffeine use in very function: impact on necrotizing enterocolitis severity. Am J Perinatol
low birth weight infants and neonatal outcomes: a systematic review and meta- 2015;32:927–32.
analysis. J Kor Med Sci 2015;30:1828–35.
379

An update on biomarkers of necrotizing enterocolitis T

∗
Pak Cheung Ng
Department of Paediatrics, Prince of Wales Hospital, The Chinese University of Hong Kong, Hong Kong, China
Keywords: Necrotizing enterocolitis (NEC) remains a devastating surgical emergency with high morbidity and mortality in
Biomarkers preterm infants. Slow but steady progress has been made in past years searching for novel biomarkers of NEC, for
Diagnosis both surveillance and diagnostic purposes. This review primarily focuses on recent discoveries: clinical appli-
Necrotizing enterocolitis cations of different categories of biomarkers for surveillance, early diagnosis, and predicting severity and
Specific
prognosis; and understanding of pathophysiological mechanisms as a basis to rationalize the search for ‘gut-
Surveillance
associated specific biomarkers’ of NEC. An important next step is to collaborate with our industrial partners to
develop point-of-care tests, and to discover novel and gut-associated specific biomarkers that can be used for
surveillance and early diagnosis of NEC in routine clinical settings.
1. Introduction more on (i) understanding of the pathophysiological mechanisms of

NEC as a basis to rationalize the search for specific biomarkers in re-
Necrotizing enterocolitis (NEC) remains one of the most frequent levant immunological pathways, (ii) highlighting recent findings on
and serious complications of prematurity. A recent survey conducted biomarkers of NEC, and (iii) more importantly, the application of dif-
between the years 2000 and 2011 on extremely premature infants in ferent categories of biomarkers in surveillance, early diagnosis, and
the USA indicated that with advances in modern diagnostic techniques predicting severity or prognosis of the condition.
and sophisticated treatments, the mortality of preterm infants asso-
ciated with respiratory diseases, infection and central nervous system 2. Categories of biomarkers in NEC
injuries had significantly improved, whereas deaths associated with
NEC were paradoxically increased [1]. In addition, NEC infants who In brief, the “non-specific” biomarkers are mediators principally of
survived developed severe morbidities such as short bowel syndrome, the proinflammatory and anti-inflammatory pathways of the im-
parenteral nutrition-associated cholestasis, poor physical growth, and munological system. These mediators are unable to differentiate NEC, a
delay in neurodevelopment [2,3]. Since the initial clinical signs and highly proinflammatory disease, from systemic neonatal sepsis, also an
symptoms of NEC, including abdominal distension, increase in gastric intensely proinflammatory state of the body, but they are very effective
aspirates or alteration of bowel habits, are relatively non-specific, and in separating NEC/sepsis from non-sepsis conditions such as apnoea of
since definitive radiological features of pneumatosis intestinalis, portal prematurity, bronchopulmonary dysplasia, benign gastrointestinal (GI)
venous gas or peritoneal free gas usually signify an advanced disease dysmotility of prematurity, electrolyte-induced or postoperative-in-
state, a quantitative biomarker that is specific in indicating bowel induced ileus [5]. This category of biomarkers includes: (i) acute phase
jury or NEC and readily available for laboratory measurement would be proteins and other proteins or lipids: C-reactive protein (CRP) [6,7],
ideal for screening and/or early diagnosis [4]. In the past two decades, serum amyloid A (SAA) [8], procalcitonin [9], inter-α-inhibitor pro-
there have been no efficient and effective “surveillance” or specific teins [10], arginine [11], apolipoprotein CII [8], platelet-activating
“diagnostic” biomarkers available for routine clinical management. factor (PAF) and related lipids [12]; (ii) cytokines: interleukin (IL)-6, IL-
In my previous review in 2014, I have discussed the pros and cons of 10, tumor necrosis factor (TNF)-α, and chemokines: IL-8, regulated on
different categories of biomarkers of NEC, namely (i) non-specific, (ii) activation normal T-cell expressed and secreted (RANTES), interferon-
enhanced non-specific, and (iii) gut-associated specific biomarkers, and γ-inducible protein-10 (IP-10); (iii) cell surface antigens: neutrophil
I have outlined a stringent protocol that may be useful as a methodo- CD64, neutrophil CD11b, and natural killer cell CD69 [7,13–17].
logical model for discovery of novel organ-specific or disease-specific The second category is also non-specific inflammatory mediators,
biomarkers for preterm infants [4]. In this review update, I shall focus calprotectin [18–20] or S100A12 [21], and these are released in
∗
Department of Paediatrics, 6th Floor, Lui Che Woo Clinical Sciences Building, Prince of Wales Hospital, Shatin, N.T., Hong Kong SAR, China.
E-mail address: pakcheungng@cuhk.edu.hk.
P.C. Ng Seminars in Fetal and Neonatal Medicine 23 (2018) 380–386
abundant quantity into the bowel lumen by neutrophils or monocytes progression of disease process, can pave the way for future discovery of
as the result of GI mucosal damage and inflammation. These com- organ-specific or disease-specific surveillance and/or diagnostic bio-
pounds are highly resistant to bacterial degradation. In addition, the markers of NEC.
composition of volatile organic compounds (VOCs) produced by colonic
fermentation is significantly altered as the result of adverse changes in 4. Clinical applications
GI microbial colonization predating and during NEC [22,23]. Though
these proteins and VOCs are by themselves not specific indicators of It is clinically relevant to discuss NEC biomarkers pertaining to
NEC, by virtue of their abnormal quantitative profiles in fecal speci- those suitable for ‘surveillance’ and those for ‘early diagnosis’ ( ± pre-
mens they would automatically point to an active inflammatory process diction of the severity of disease or prognosis), since different types of
developing in the GI tract. Thus, they are referred to as the “enhanced biological specimens and biomarkers possess different physical and
non-specific” biomarkers of NEC. biochemical properties.
More importantly, the third category comprises the “gut-associated
specific biomarkers,” which include fatty acid binding proteins (FABP), 4.1. Surveillance
trefoil factor-3 (TFF-3), claudin-3, and others [24,25]. These proteins
are components of the GI mucosa and bowel wall. Upon damage in- Genetic predisposing factors of racial and gender disparities, for
duced by inflammation, hypoxic–ischemic insults and/or bacterial in- example, black male infants, were reported to have increased risks of
vasion, these GI-specific proteins are released in large quantities from developing NEC [32], and genetic variations in single nucleotide
the damaged bowel into the bloodstream [25] and are subsequently polymorphisms (SNPs), for example, NfκB1 or NfκB1A [33] and others,
excreted in the urine [24]. Thus, these biomarkers are specific for di- have been implicated in the condition. The presence of aberrantly high
agnosing GI injuries independent of systemic neonatal sepsis [25] and concentrations of oxidative stress mediators in umbilical cord blood,
are most useful clinically in determining the exact course of manage- such as hydroperoxides or oxidation protein products, in the perinatal
ment such as broadening the spectrum of antibiotics to cover anaerobic period have also been suggested as risk factors [34]. Further, Cordeiro
gut organisms, early initiation of total parenteral nutrition, and close et al. claimed to have developed a novel mathematical model of 27 cord
radiological monitoring for adverse abdominal complications in gen- blood biomarkers consisting mainly of proinflammatory, anti-in-
uine NEC cases. Therefore, understanding the biochemical properties of flammatory and chemotactic cytokines and soluble receptors, neuro-
the mediators, nature of the specimens, pathophysiological pathways/ tropins, and matrix metalloproteases, for predicting adverse neonatal
mechanisms, and limitations of different categories of biomarkers, outcomes of intracranial hemorrhage, sepsis and NEC in preterm infants
frontline neonatologists may be able to use the most appropriate [35]. More recently, a retrospective cohort study using metabolic pro-
mediators and specimen mediums for surveillance and diagnostic pur- filing through newborn screening has shown that acylcarnitine profiles
poses. could reflect an increased risk of NEC [36]. Each log unit rise in C5 and
free carnitine/(C16 + 18:1) would be correlated with a 78% and 76%
3. Genome-wide expression profiles and pathophysiological increased risk for developing the disease [36]. In addition, five weekly
pathways of human NEC whole-blood specimens collected from 177 infants born ≤32 weeks
gestation and/or ≤1500 g for measurement of reticulated platelets (RP)
Our research team has recently investigated the molecular me- and intestinal alkaline phosphatase (iAP) revealed that, of 15 infants
chanisms and constructed a comprehensive database on differential who subsequently developed NEC, 14 (93%) had low (≤2.3%) RP and
gene expression profiles of human bowel tissues from NEC and spon- nine (60%) had high iAP (> 0 U/L) in at least one specimen before the
taneous intestinal perforation (SIP) cases [26]. We demonstrated ex- onset of disease [37]. Others have suggested that the identification of
tensive dysregulation of gene expression in NEC but not SIP, comprising microbial signatures in GI microbiota, for example, clostridium and
multiple functional pathways of inflammation, hypoxia and oxidative klebsiella operational taxonomic units (OTUs) in fecal samples, could
stress, chemotaxis, cell adhesion, arginine metabolism, angiogenesis, precede NEC diagnosis in preterm infants [38]. In my opinion, all these
extracellular matrix remodeling, and muscle contraction (Fig. 1). These aforementioned potential risk indicators could be synergistically com-
dysregulated genes can be networked downstream of key receptors, bined with neonatal clinical factors, such as extreme prematurity, for-
including toll-like receptor (TLR)-2, TLR-4, CD-14 and triggering re- mula-milk feeding, prolonged antibiotics or mechanical ventilation
ceptor expressed on myeloid cell 1 (TREM 1); their signals amplify via usage, and frequent requirement of red cell transfusions, to formulate a
major transcription factors nuclear factor kappa B (NFΚB), activator risk-weighted statistical model for identifying those preterm infants
protein-1 (AP-1) and CAMP responsive element binding protein-1 most at risk of NEC for targeted surveillance [39].
(CREB-1), which further orchestrate chained activities resulting in Continuous monitoring of physiological parameters in preterm in-
chemotaxis and exaggerated microbial-mediated and/or non-microbial- fants has been advocated as a non-invasive tool for early and rapid
mediated proinflammatory responses. Clinically, markedly upregulated recognition of sepsis/NEC [40,41]. Atypical electrocardiographical
proinflammatory mediators, including cytokines, chemokines and po- patterns of decreased heart rate variability or transient deceleration
tent phospholipids (e.g. PAF [27]), cause widespread epithelial cell may be translated into a clinical score for easy quantification of the
damage and disruption of cellular tight junctions, thereby increasing frequency and magnitude of these abnormal discharges [41]. In future,
mucosal permeability, inflammatory cell infiltration (e.g. monocytes) such non-invasive monitoring systems could be used to alert frontline
and aggregation (e.g. leukocytes and platelets) in inflammatory sites, neonatologists for early performing of gut-associated specific biomarker
vasoconstriction and ultimately resulting in severe tissue damage and tests or sepsis work-up with immediate commencement of antibiotic
cell death due to necrosis and apoptosis [26,27]. The proinflammatory treatment.
influences of micro(mi)-RNA have also been explored in a similar but Recently, our investigation team has explored the use of neutrophil
separate study (Fig. 2) [28]. These informational platforms of mRNA CD64 as a non-specific screening biomarker for daily surveillance of
and miRNA array data are publicly available at Gene Express Omnibus neonatal sepsis and NEC [16]. Daily surveillance was feasible because
(GEO) database (Assession number: GSE46619 and GSE68054) [26,28]. laboratory analysis using flow cytometry required only a very tiny
Thus, understanding the underlying molecular genetics, biochemical/ volume of whole blood (0.05 mL which is equivalent to one or two
metabolic (e.g. gluconate) [29] and hematological engagement (e.g. drops of blood) and we have also confirmed that there was excellent
migration of monocytes [30], and alteration of FOXP3 T regulatory cell quantitative agreement of neutrophil CD64 expression between paired
(Treg): CD4 T-cell ratio or Treg:CD8 T-cell ratio in damaged bowel venous and capillary blood samples (n = 33 pairs, r = 0.999) [16]. In
tissues [31]) in various pathways during the development and this surveillance study, a total of 146 infants were consecutively studied
381
Fig. 1. Pathway traced network of dysregulated target genes in necrotizing enterocolitis (NEC). Quantitative polymerase chain reaction-validated genes that were
found to be associated with NEC [26] were used to construct a biological network based on the MetaCore software using a pathway traced algorithm. Expression
changes are shown as red and blue circles, representing up- and downregulation, respectively. Green, red, and grey arrows between target genes represent positive,
negative, and unspecified interactions, respectively. Functional categories, including angiogenesis, arginine metabolism, cell adhesion, chemotaxis and inflammation,
extracellular matrix remodeling, hypoxia/oxidative stress, and muscle contraction are highlighted in different colors (Adopted from Ref. [26] with copyright per-
mission obtained from Annals of Surgery, license number: 4278111403743.).
and 155 episodes of sepsis evaluations were performed. The screening assess the number of unnecessary sepsis work-ups required to be per-
utilities comprising sensitivity, specificity, positive and negative pre- formed, thus defeating the main purpose of asymptomatic screening or
dictive values for surveillance of late-onset sepsis/NEC were 89%, 98%, surveillance.
41% and 99.8%, respectively [16]. Genuine sepsis/NEC episodes were Blood is never an ideal specimen for surveillance purposes because
detected on average 1.5 days before clinical presentation. However, 63 blood-taking procedures are unpleasant and invasive. Biomarkers pre-
episodes of CD64 activation occurred above the cut-off CD64 expression sent in other specimen mediums which could be collected non-in-
level (5655 antibody-phycoerythrin (PE) molecules bound/cell) in vasively, such as stool, urine or buccal swab, are considered more ap-
asymptomatic infants who did not require antibiotic treatment and who propriate. Stool as a screening specimen medium has been extensively
would not otherwise required sepsis work-up. It was likely that these studied. Non-specific proteins, including calprotectin [18–20] and
episodes represented the host response to either minor viral infection or S100A12 [21], and more recently VOC profiles [22,23], have been the
transient bacterial translocation which could be eradicated by the in- focus of investigation for screening of NEC. Fecal calprotectin and
fants' own intrinsic immunological defense. These episodes were defi- S100A12 are “enhanced non-specific” biomarkers that possess similar
nitely not spurious laboratory measurements, as CD64 levels rose and clinical and biochemical properties. The presence of high concentra-
declined gradually over a period of four to five days [16]. Such sur- tions of these proteins in fecal samples signifies disruption of mucosal
veillance results adversely translated into an additional 41% un- integrity and increased intestinal permeability due to bowel wall injury
necessary sepsis work-ups to be performed if the surveillance criteria of and/or inflammation. Disappointingly, the wide intra- and inter-in-
pre-symptomatic sepsis evaluation and antibiotics treatment were to be dividual variations of concentrations in stool, significant influence by
routinely adopted and strictly followed [16]. Two other surveillance gestational and postnatal age, substantial overlapping of concentrations
studies using IL-6 and IL-1 receptor antagonist (IL-1ra) [42], and neu- between the normal and NEC group, and lack of optimal cut-off values
trophil CD11b [17] were performed in smaller preterm cohorts for defining an abnormal concentration in fecal samples, rendered these
(n = 101 and n = 30, respectively). IL-6 and IL-1ra were found to be fecal protein results extremely difficult to interpret [18–21]. Despite
upregulated two days prior to clinical diagnosis of sepsis [42], and claims of significantly earlier detection (four to 10 days before clinical
neutrophil CD11b gradually increased over a three-day period before diagnosis of NEC) [21] and significantly higher calprotectin and
sepsis work-ups were performed [17]. However, as each daily level of S100A12 fecal concentrations in NEC compared with non-NEC infants
these mediators should be considered as an independent indicator for in some studies, these proteins have not been recommended for routine
performing pre-symptomatic sepsis evaluation, these two latter studies clinical surveillance in most neonatal centers. However, rapid fecal
did not address this important issue [17,42] and so we were unable to calprotectin assay as a point-of-care test has been recently developed
382
Fig. 2. Pathway traced network of miRNA/mRNA pairs in necrotizing enterocolitis (NEC). A functional network on the association of selected miRNAs and potential
target mRNAs was generated using the MetaCore software and sequence-based prediction. Quantitative polymerase chain reaction-validated expression changes of
miRNAs or mRNAs are shown as red and blue circles, representing up- and downregulation, respectively. Green, red, and grey arrows between target genes represent
positive, negative, and unspecified interactions, respectively. Large arrows between miRNA/mRNA pairs indicate experimentally proven relationship. Transcription
factors and functional categories, including angiogenesis, arginine metabolism, cell adhesion, chemotaxis and inflammation, extracellular matrix remodeling, hy-
poxia/oxidative stress, and muscle contraction are highlighted in different colors. miRNAs exhibiting significant inverse correlation with specific mRNAs are un-
derlined. (Adopted from Ref. [28] with copyright permission automatically granted by the open access journal PLoS One.).
and was assessed in a pilot study to have good agreement (r = 0.899) NEC would not pass stool for a prolonged period as a result of severe
with the conventional enzyme-linked immunosorbent fecal calprotectin ileus [4]. Thus, fecal sample as a non-invasive specimen medium would
assay (13 NEC versus 16 non-NEC stool samples) [43]. If this technique probably not be ideal for both surveillance or acute diagnostic pur-
could be subsequently proven valid and reliable, a point-of-care rapid poses.
screening test would be considered a valuable tool for surveillance and/ In a proof-of-concept proteomic protein discovery study [44], cu-
or early identification of NEC. A recent case–control study using daily mulative buccal swab specimens were collected from very low birth
stool sample for measuring altered VOC profiles via the eNose tech- weight infants at one, two and three weeks before the development of
nique demonstrated good sensitivity and specificity (83% and 75%, NEC (n = 10) and in matched control infants (n = 10). The preliminary
respectively) for discriminating NEC cases from controls, two to three result from samples collected two to three weeks before development of
days before the onset of clinical NEC [23]. Such encouraging results are NEC indicated a lower IL-1ra concentration compared with non-NEC
required to be confirmed in a larger prospective cohort study. As infants. Buccal swab as a non-invasive specimen medium might in fu-
mentioned in my previous review [4], enhanced non-specific stool ture provide an alternative choice for surveillance of NEC.
biomarkers have major intrinsic limitations. Since fecal samples are
usually paste-like or in a solid medium in newborn infants and focal 4.2. Early diagnosis
bowel inflammation would produce proinflammatory proteins that are
excreted into its adjacent lumen, these protein mediators could not be Since surveillance studies are notoriously difficult to perform be-
homogeneously dispersed throughout the entire specimen. It was also cause specimens are required to be longitudinally collected and ana-
our and others' experience that a significant proportion of infants with lysed either on a daily basis or at frequent regular intervals, most
383
studies focused on early diagnosis after the onset of signs and symptoms of urine proteins also found a panel of seven mediators (α-2-macro-
or assessing the severity of NEC with an objective to accurately dif- globulin-like protein 1, cluster of differentiation protein 14, cystatin 3,
ferentiate severe surgical NEC from milder medical cases. fibrinogen α-chain, pigment epithelium-derived factor, retinal binding
Recently, research interest has been directed to evaluate hemato- protein 4, and vasolin) which could potentially differentiate NEC from
logical parameters as biomarkers of NEC. Based on the hypothesis that sepsis cases, as well as surgical NEC from medical cases with good ac-
circulating monocytes are recruited to the inflamed bowel due to che- curacy [52]. Our research team has spent the past five years planning
motaxis, a significant fall in circulating absolute monocyte count (AMC) and conducted a study on micro (mi)-RNA for discovering novel gut-
could be potentially used as a warning indicator [30]. In a retrospective associated specific biomarkers for diagnosis of both mild medical and
study comparing 69 NEC cases with 257 matched controls having be- severe surgical NEC. Using (i) the microarray method to differentiate
nign feeding intolerance but without signs of NEC, a significant fall in various mi-RNAs among cases in the NEC, sepsis and non-NEC/non-
AMCs from medians 1.7 × 109/L and 2.1 × 109/L to 0.8 × 109/L and sepsis groups, (ii) selecting potential target mi-RNAs using a case–-
0.8 × 109/L in stages II and III NEC, respectively, were observed. There control study with matched “healthy” controls, and (iii) ultimately
was no change in AMCs in matched control infants. The investigators testing the selected promising mi-RNAs in a large cohort study, we
suggested that a fall in AMCs > 20% could indicate NEC in infants with identified three novel mi-RNAs which could specifically distinguish
feeding intolerance [30]. Although the sensitivity (0.70; 95% con- both medical and surgical NEC from sepsis and non-NEC/non-sepsis
fidence intervals (CI): 0.57–0.81) and specificity (0.71; 95% CI: cases with very favorable diagnostic utilities and accuracy (unpublished
0.64–0.77) of AMCs were only modest, this parameter would be easily data). Our study illustrated that researchers now possess very powerful
and readily available from all hospital laboratories [30]. In another investigative tools for proceeding to identify novel biomarkers for
recent study, intestinal epithelial damage would trigger eosinophilic specific organ-injury and/or specific disease.
activation resulting in eosinophilia in NEC [45]. Persistent eosinophila
≥5% for ≥5 days was claimed to have eight-fold increased risk for 4.3. Assessing severity
developing serious complications, especially hepatic fibrosis or in-
testinal strictures, during NEC convalescence [45]. In addition, Tao Many studies reported the assessment of disease severity in con-
et al. in a case–control study comparing preterm infants with NEC junction with biomarker discovery but only very few were purposefully
(n = 40; 20 medical NEC and 20 surgical NEC cases), sepsis (n = 20) designed solely to address this issue. As a general rule, the greater the
and healthy controls (n = 24) observed significant impairment in the deviation of biomarkers from their baseline levels, the more severe
formation of fibrinogen-γ -dimer (FGG-dimer) in the NEC group during would be the disease severity. The degree of deviation would reflect the
plasma protein profiling [46]. The replenishment of Factor XIII into magnitude of activation of these mediators in the immunological
NEC plasma could correct this defect. It was postulated that FGG-dimer pathways. Further, sustained aberrant levels of biomarkers could be due
could effectively differentiate NEC from sepsis or healthy controls and to persistent dysregulation of critical pathophysiological pathways,
could be potentially developed as a novel specific biomarker [46]. indicating a higher chance of developing complications and adverse
Currently, the most widely used biomarkers of NEC are still the non- outcomes. The use of multiple biomarkers (or combined with electro-
specific biomarkers of acute phase reactants, for example, CRP, SAA or physiological parameters) would constitute a clearer clinical status,
the ApoSAA score (combining SAA and apolipoprotein CII) [6,8,9,13]. since information obtained from physiological, hematological and re-
A newly advocated enzyme, β-glucosidase, though claimed to have levant biochemical pathways could be aggregated to reveal an unbiased
sensitivity 84.6% and specificity 85.9% for identifying NEC infants, has clinical picture.
a disappointingly low positive predictive value of 37.9%, rendering it We have previously reported that the sequential use of IL-10, IL-6
unfavorable to be considered as a clinically useful diagnostic marker and RANTES could accurately predict the development of disseminated
[47]. Cytokines and chemokines, for example, IL-6, IL-8, IL-10 and IP- intravascular coagulation (DIC) in infants with systemic sepsis and NEC
10 [13–15], though proven to be effective ‘early warning’ biomarkers of at the onset of clinical presentation with 100% sensitivity and 97%
sepsis/NEC, have not been used widely in real-life clinical settings. The specificity [15]. The evaluation of proinflammatory:anti-inflammatory
requirement for laboratory analysis in batches and high costs probably cytokine ratio(e.g. IL-6:IL-10) could reflect the severity of disease and
limited their usefulness as routine diagnostic tools. Cell surface antigens accurately predict mortality [15,53].
with neutrophil CD64, in particular, was an excellent early-warning Recent studies have attempted to identify severe NEC cases which
biomarker of sepsis/NEC and was considered superior than CRP for required surgical intervention so that early transfer to tertiary pediatric
diagnosing intra-abdominal catastrophe [7]. However, the lack of au- surgical centers and treatment could be efficiently initiated. High IL-8
tomated method of measurement has greatly confined its use as a re- levels with a cut-off level > 1783 pg/mL at diagnosis have been asso-
search tool and in well-equipped tertiary centers with flow cytometric ciated with NEC infants who subsequently required surgery (sensitivity
facilities. 90.5% and specificity 59.2%) [51]. Similarly, urinary SAA levels were
With advances in sequencing and omics technologies, neonatolo- significantly higher in surgical NEC compared with medical NEC infants
gists should devote their efforts to discover novel gut-associated specific (cut-off level > 34.4 ng/mL yielding sensitivity 83% and specificity
biomarkers for differentiating NEC from neonatal sepsis or other sys- 83%) [54]. Combining urinary SAA with platelet counts further in-
temic proinflammatory conditions. As described in detail in my pre- creased the sensitivity to 94% without affecting the specificity [54].
vious review [4], both blood or urine intestinal (I)-FABP, liver (L)- The panel of seven urinary biomarkers previously described has also
FABP, claudin-3 and TFF-3 concentrations have been shown to be been claimed to be capable of separating surgical NEC from medical
useful indicators for specific diagnosis of acute NEC [24,25,48,49]. NEC cases with high precision [52]. Yakut et al. further described the
Combining plasma L-FABP, I-FABP and TFF-3 to formulate the LIT score use of serial ischemia-modified albumin (IMA) and compared it with
could identify severe surgical NEC or death cases versus uncomplicated CRP and IL-6. IMA levels were significantly higher in infants with stage
medical NEC cases with high precision [25]. Serial I-FABP measure- III NEC than those with stage II disease, and IMA was better in pre-
ments also predicted the development of complicated disease [48]. A dicting perforation and mortality [55]. Since severe NEC frequently
recent study showed that plasma I-FABP concentrations correlated causes tissue damage extending into the intestinal muscularis layer of
strongly with urine I-FABP levels, thus offering an opportunity to the gut, smooth muscle actin (SMA) has been postulated to be released
choose the most appropriate specimen for measurement [50]. Further, in significant quantities into the bloodstream [56]. A pilot study re-
calculating urinary I-FABP/creatinine ratio appeared redundant [50]. vealed that SMA was detected in plasma from all infants with severe
Not unexpectedly, FABPs also correlated significantly with non-specific surgical NEC (n = 4) but not in other infants with focal disease or those
biomarkers IL-6 [50] and IL-8 [51]. An exploratory proteomics analysis who were successfully treated medically [56].
384
Sho et al. attempted to develop a scoring system aiming to identify 5.1. Practice points
infants at risk of developing the most severe form of NEC-totalis from its
milder variants upon the onset of clinical presentation [57]. Older • There has been a slow but steady progress in the past decade
postnatal age at diagnosis, low platelet counts, and high phosphorus searching for novel biomarkers of NEC. Most currently used clinical
and creatinine levels were found to be significant adverse factors. A 0–5 biomarkers are “non-specific” mediators for early diagnosis.
point scale “Totalis Score” was developed and evaluated among 157 • Novel gut-associated specific plasma mi-RNA biomarkers for iden-
NEC infants with 13 having NEC-totalis. This scoring system was found tification of both mild medical and severe surgical NEC cases have
to have sensitivity 92% and specificity 78% for predicting this most been recently discovered (unpublished data).
severe form of NEC [57]. Similarly, we have derived the 0–9 point scale
LIT score based on three gut-associated specific biomarkers (L-FABP, I- 5.2. Research directions
FABP and TFF-3) [25]. With a median cut-off of 4.5 points, the LIT
score identified surgical NEC with sensitivity and specificity of 83% and • Further research must be directed to discover a reliable “gut-asso-
100%, respectively. Also, the LIT score was significantly higher in NEC ciated specific biomarker” for accurate and early diagnosis of NEC.
cases with DIC than among those without DIC [25]. In future, I would • Investigators should collaborate with industrial partners to develop
anticipate that such scoring systems or algorithms combining key a practical “point-of-care test” which may allow frontline neona-
clinical parameters plus different classes of laboratory biomarkers tologists to screen for probable NEC cases at the bedside.
would provide more sophisticated information than a single biomarker • Theoretically, sophisticated algorithms or complex scores com-
test, and would dominate the clinical diagnostic trend for genuinely prising clinical, laboratory and radiological data can be downloaded
reflecting the “probability” of the infant having NEC or a severe variant to a mobile electronic device for calculating the “probability” of a
of the disease. By simply entering key clinical and laboratory data into a particular infant having NEC.
computer or mobile electronic device, even complex clinical algorithms
could be easily handled to quantify the “risk” for an accurate diagnosis Conflicts of interest
or assessment of disease severity [5].
None declared.
5. Conclusion
Funding source
In contrast to my previous review of NEC biomarkers [4] that
mainly focused on the biochemical characteristics of different cate- The NEC research was supported by a donation from the charity
gories of NEC mediators, the current article has concentrated more on organization Providence Foundation Limited, Hong Kong (Project code:
clinical use at various clinical dimensions, including daily surveillance, 6901814). The funding organization was not involved in the study
acute diagnosis, and predicting severity and prognosis of the disease. design, data collection, data analysis, decision to publish or preparation
Currently, the mainstay of biomarker usage is for early identification of of the manuscript.
NEC using non-specific inflammatory biomarkers such as CRP, SAA, or
neutrophil CD64. With increasing understanding of genome-wide ex- References
pression profiles and pathophysiological pathways using high-
throughput arrays and omic technologies, gut-associated specific bio- [1] Patel RM, Kandefer S, Walsh MC, et al. Causes and timing of death in extremely
markers have emerged as a superior class of mediators that can speci- premature infants from 2000 through 2011. N Engl J Med 2015;372:331–40.
[2] Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011;364:255–64.
fically pinpoint injuries to a single organ. A valid surveillance study is [3] Salhab WA, Perlman JM, Silver L, Sue Broyles R. Necrotizing enterocolitis and
notoriously difficult to perform and to comply with the follow-up in- neurodevelopmental outcome in extremely low birth weight infants < 1000 g.
vestigations and therapeutic actions in real-life clinical settings. Such J Perinatol 2004;24:534–40.
[4] Ng PC. Biomarkers of necrotising enterocolitis. Semin Fetal Neonatal Med
tests must be non-invasive and the nature of specimens carefully 2014;19:33–8.
chosen. Importantly, for a surveillance program to be successful, [5] Ng PC, Ma TPY, Lam HS. The use of laboratory biomarkers for surveillance, diag-
frontline neonatologists must have the confidence to act in response to nosis and prediction of clinical outcomes in neonatal sepsis and necrotising en-
terocolitis. Arch Dis Child Fetal Neonatal Ed 2015;100:F448–52.
aberrant laboratory results which may predate the presentation of [6] Pourcyrous M, Korones SB, Yang W, Boulden TF, Bada HS. C-reactive protein in the
clinical signs in asymptomatic infants [16]. Inevitably, some un- diagnosis, management and prognosis of neonatal necrotizing enterocolitis.
necessary investigations and/or treatments would be performed, but if Pediatrics 2005;116:1064–9.
[7] Lam HS, Wong SP, Cheung HM, Chu WC, Wong RP, Chui KM, et al. Early diagnosis
the surveillance test has good positive predictive value, the number of
of intra-abdominal inflammation and sepsis by neutrophil CD64 expression in
false-positive cases should be kept to a minimum. Understandably, the newborns. Neonatology 2011;99:118–24.
severity of disease is often reflected by the magnitude of dysregulation [8] Ng PC, Ang IL, Chiu RW, Li K, Lam HS, Wong RP, et al. Host-response biomarkers
in their biochemical pathways. The balance between pro-inflammatory for diagnosis of late-onset septicemia and necrotizing enterocolitis in preterm in-
fants. J Clin Invest 2010;120:2989–3000.
and anti-inflammatory responses will give accurate indications for [9] Centinkaya M, Ozkan H, Koksal N, Akaci O, Ozgür T. Comparison of the efficacy of
predicting DIC and mortality [15,53]. serum amyloid A, C-reactive protein, and procalcitonin in the diagnosis and follow-
Overall, steady progress has been achieved in the field of biomarker up of necrotizing enterocolitis in premature infants. J Pediatr Surg 2011;46:1482–9.
[10] Chaaban H, Shin M, Sirya E, Lim YP, Caplan M, Padbury JF. Inter-alpha inhibitor
research for NEC in recent decades. One of the main difficulties of why protein level in neonates predicts necrotizing enterocolitis. J Pediatr
so many promising biomarkers could not find their way into routine 2010;157:757–61.
clinical practice was due to the lack of industrial collaboration for de- [11] Richir MC, Siroen MP, van Elburg RM, et al. Low plasma concentrations of arginine
and asymmetric dimethylarginine in premature infants with necrotizing en-
veloping the laboratory analysis into a fully automated test, thereby terocolitis. Br J Nutr 2007;97:906–11.
transforming them to be user-friendly and cost-effective. More im- [12] Rabinowitz SS, Dzakpasu P, Piecuch S, Leblanc P, Valencia G, Kornecki E. Platelet-
portantly, I can foresee future advances in three major areas: (i) dis- activating factor in infants at risk for necrotizing enterocolitis. J Pediatr
2001;138:81–6.
covery of more novel and specific biomarkers; (ii) conversion of com- [13] Ng PC, Cheng SH, Chui KM, et al. Diagnosis of late onset neonatal sepsis with cy-
plex laboratory measurements into simple point-of-care bedside tokines, adhesion molecule, and C-reactive protein in preterm very low birth weight
diagnostic tests; and (iii) use of sophisticated algorithms via mobile infants. Archs Dis Childh Fetal Neonatal Ed 1997;77:221–7.
[14] Ng PC, Li K, Chui KM, et al. IP-10 is an early diagnostic marker for identification of
electronic devices either as an app or a computer program to accurately
late-onset bacterial infection in preterm infants. Pediatr Res 2007;61:93–8.
calculate the “probability” of having the disease or assessing the se- [15] Ng PC, Li K, Leung TF, et al. Early prediction of sepsis-induced disseminated in-
verity of NEC, for example, surgical versus medical disease, and mor- travascular coagulation with interleukin-10, interleukin-6, and RANTES in preterm
tality of this devastating condition in preterm infants. infants. Clin Chem 2006;52:1181–9.
385
[16] Lam HS, Cheung HM, Poon TCW, et al. Neutrophil CD64 for daily surveillance of 2017;181:80–5.
systemic infection and necrotizing enterocolitis in preterm infants. Clin Chem [37] Kampanatkosol R, Thomson T, Habeeb O, et al. The relationship between re-
2013;59:1753–60. ticulated platelets, intestinal alkaline phosphatase, and necrotizing enterocolitis.
[17] Turunen R, Andersson S, Nupponen I, Kautiainen H, Siitonen S, Repo H. Increased J Pediatr Surg 2014;49:273–6.
CD11b-density on circulating phagocytes as an early sign of late onset sepsis in [38] Sim K, Shaw AG, Randell P, et al. Dysbiosis anticipating necrotizing enterocolitis in
extremely low-birth-weight infants. Pediatr Res 2005;57:270–5. very premature infants. Clin Infect Dis 2015;60:389–97.
[18] Josefsson S, Bunn SK, Domellof M. Fecal calprotectin in very low birth weight in- [39] Markel TA, Engelstad H, Poindexter BB. Predicting disease severity of necrotizing
fants. J Pediatr Gastroenterol Nutr 2007;44:407–13. enterocolitis: how to identify infants for future novel therapies. J Clin Neonatol
[19] Zoppelli L, Guttel C, Bittrich HJ, Andree C, Wirth S, Jenke A. Fecal calprotectin 2014;3:1–9.
concentrations in premature infants have a lower limit and show postnatal and [40] Stone ML, Tatum PM, Weitkamp JH, et al. Abnormal heart rate characteristics
gestational age dependence. Neonatology 2012;102:68–74. before clinical diagnosis of necrotizing enterocolitis. J Perinatol 2013;33:847–50.
[20] Selimoglu MA, Temel I, Yildirim C, Ozyalin F, Aktas M, Karabiber H. The role of [41] Fairchild KD. Predictive monitoring for early detection of sepsis in neonatal ICU
fecal calprotectin and lactoferrin in the diagnosis of necrotizing enterocolitis. patients. Curr Opin Pediatr 2013;25:172–9.
Pediatr Crit Care Med 2012;13:452–4. [42] Küster H, Weiss M, Willeitner AE, et al. Interleukin-1 receptor antagonist and in-
[21] Dabritz J, Jenke A, Wirth S, Foell D. Fecal phagocyte-specific S100A12 for diag- terleukin-6 for early diagnosis of neonatal sepsis 2 days before clinical manifesta-
nosing necrotizing enterocolitis. J Pediatr 2012;162:1059e64. tion. Lancet 1998;352:1271–7.
[22] Garner CE, Ewer AK, Elasouad K, et al. Analysis of faecal volatile organic com- [43] Bin-Nun A, Booms C, Sabag N, Mevorach R, Algur N, Hammerman C. Rapid fecal
pounds in preterm infants who develop necrotizing enterocolitis: a pilot study. calprotectin (FC) analysis: point of care testing for diagnosing early necrotizing
J Pediatr Gastroenterol Nutr 2009;49:559–65. enterocolitis vol. 32. 2015. p. 337–42.
[23] de Meij TGJ, van der Schee MPC, Berkhout DJC, et al. Early detection of necrotizing [44] Torrazza RM, Li N, Young C, et al. Pilot study using proteomics to identify pre-
enterocolitis by fecal volatile organic compounds analysis. J Pediatr dictive biomarkers of necrotizing enterocolitis from buccal swabs in very low birth
2015;167:562–7. weight infants. Neonatology 2013;104:234–42.
[24] Thuijls G, Derikx JP, van Wijck K, et al. Non-invasive markers for early diagnosis [45] Wahidi LS, Sherman J, Miller MM, Zaghouani H, Sherman MP. Early persistent
and determination of the severity of necrotizing enterocolitis. Ann Surg blood eosinophilia in necrotizing enterocolitis is a predictor of late complications.
2010;251:1174–80. Neonatology 2015;108:137–42.
[25] Ng EW, Poon TC, Lam HS, et al. Gut-associated biomarkers L-FABP, I-FABP, and [46] Tao GZ, Liu B, Zhang R, et al. Impaired activity of blood coagulant Factor XIII in
TFF3 and LIT score for diagnosis of surgical necrotizing enterocolitis in preterm patients with necrotizing enterocolitis. Sci Rep 2015;5:13119.
infants. Ann Surg 2013;258:1111–8. [47] Moreno JLG-C, Torronteras AR, Gonzalez MDR, et al. The β-glucosidase assay: a
[26] Chan KY, Leung KT, Tam YH, et al. Genome-wide expression profiles of necrotizing new diagnostic tool for necrotizing enterocolitis. Sensitivity, specificity, and pre-
enterocolitis versus spontaneous intestinal perforation in human intestinal issue: dictive values. Eur J Pediatr 2016;175:931–41.
dysregulation of functional pathways. Ann Surg 2014;260:1128–37. [48] Schurink M, Kooi EMW, Hulzebos CV, et al. Intestinal fatty acid-binding protein as a
[27] Frost BL, Caplan MS. Necrotizing enterocolitis: pathophysiology, platelet-activating diagnostic marker for complicated and uncomplicated necrotizing enterocolitis: a
factor, and probiotics. Semin Pediatr Surg 2013;22:88–93. prospective cohort study. PLoS One 2015;10:e0121336.
[28] Ng PC, Chan KY, Leung KT, et al. Comparative miRNA expressional profiles and [49] Benkoe TM, Mechtler TP, Weninger M, Pones M, Rebhandl W, Kasper DC. Serum
molecular networks in human small bowel tissues of necrotizing enterocolitis and levels of interleukin-8 and gut-associated biomarkers in diagnosing necrotizing
spontaneous intestinal perforation. PLoS One 2015;10:e0135737. enterocolitis in preterm infants. J Pediatr Surg 2014;49:1446–51.
[29] Dessi A, Pintus R, Marras S, Marincola FC, De Magistris A, Fanos V. Metabolomics in [50] Schurink M, Scholten IGH, Kooi EMW, Hulzebos CV, Kox RG, Groen H. Intestinal
necrotizing enterocolitis: the state of the art. Expert Rev Mol Diagn fatty acid-binding protein in neonates with imminent necrotizing enterocolitis.
2016;16:1053–8. Neonatology 2014;106:49–54.
[30] Remon J, Kampanatkosol R, Kaul RR, Muraskas JK, Christensen RD, Maheshwari A. [51] Benkoe T, Reck C, Pones M, Weninger M, Gleiss A, Stift A. Interleukin-8 predicts 60-
Acute drop in blood monocyte count differentiates NEC from other causes of feeding day mortality in premature infants with necrotizing enterocolitis. J Pediatr Surg
intolerance. J Perinatol 2014;34:549–54. 2014;49:385–9.
[31] Weitkamp JH, Koyama T, Rock MT, et al. Necrotising enterocolitis is characterised [52] Sylvester KG, Ling XB, Liu GY, et al. Urine protein biomarkers for the diagnosis and
by disrupted immune regulation and diminished mucosal regulatory (FOXP3)/ef- prognosis of necrotizing enterocolitis in infants. J Pediatr 2014;164:607–12.
fector (CD4, CD8) T cell ratios. Gut 2013;62:73–82. [53] Ng PC, Li K, Wong RP, et al. Proinflammatory and anti-inflammatory cytokine re-
[32] Uauy RD, Fanaroff AA, Korones SB, Phillips EA, Phillips JB, Wright LL. Necrotizing sponses in preterm infants with systemic infections. Archs Dis Childh Fetal Neonatal
enterocolitis in very low birth weight infants: biodemographic and clinical corre- Ed 2003;88:209–13.
lates. J Pediatr 1991;119:630–8. [54] Reisinger KW, Kramer BW, Van der Zee DC, et al. Non-invasive serum amyloid A
[33] Sampath V, Le M, Lane L, et al. The NFKB1 (g.-24519delATTG) variant is associated (SAA) measurement and plasma platelets for accurate prediction of surgical inter-
with necrotizing enterocolitis (NEC) in premature infants. J Surg Res vention in severe necrotizing enterocolitis (NEC). PLoS One 2014;9:e90834.
2011;169:e51–7. [55] Yakut I, Tayman C, Oztekin O, Namuslu M, Karaca F, Kosus A. Ischemia-modified
[34] Perrone S, Tataranno ML, Negro S, et al. May oxidative stress biomarkers in cord albumin may be a novel marker for the diagnosis and follow-up of necrotizing
blood predict the occurrence of necrotizing enterocolitis in preterm infants? enterocolitis. J Clin Lab Anal 2014;28:170–7.
J Matern Fetal Neonatal Med 2012;25(Suppl 1):128–31. [56] Evennett N, Cerigioni E, Hall NJ, Pierro A, Eaton S. Smooth muscle actin as a novel
[35] Cordeiro CN, Savva Y, Vaidya D, et al. Mathematical modeling of the biomarker serologic marker of severe intestinal damage in rat intestinal ischemia-reperfusion
milieu to characterize preterm birth and predict adverse neonatal outcomes. Am J and human necrotising enterocolitis. J Surg Res 2014;191:323–30.
Reprod Immunol 2016;75:594–601. [57] Sho S, Neal MD, Sperry J, Hackam DJ. A novel scoring system to predict the de-
[36] Sylvester KG, Kastenberg ZJ, Moss RL, et al. Acylcarnitine profiles reflect metabolic velopment of necrotizing enterocolitis totalis in premature infants. J Pediatr Surg
vulnerability of necrotizing enterocolitis in newborns born premature. J Pediatr 2014;49:1053–6.
386

Genetic predisposition to necrotizing enterocolitis in premature infants: T

Current knowledge, challenges, and future directions
Alain Cuna, Lovya George, Venkatesh Sampath∗
Division of Neonatology, Department of Pediatrics, Children's Mercy Kansas City, Kansas City, MO, USA
ARTICLE INFO ABSTRACT
Keywords: The role of genetics in the pathogenesis of necrotizing enterocolitis (NEC) was initially informed by epide-
Genetic predisposition miological data indicating differences in prevalence among different ethnic groups as well as concordance in
Necrotizing enterocolitis twins. These early observations, together with major advances in genomic research, paved the way for studies
Preterm infants that begin to reveal the contribution of genetics to NEC. Using the candidate gene or pathway approach, several
Single nucleotide polymorphisms
potential pathogenic variants for NEC in premature infants have already been identified. More recently, genome-
Toll-like receptors
wide association studies and exome-sequencing based studies for NEC have been reported. These advances,
however, are tempered by the lack of adequately powered replication cohorts to validate the accuracy of these
discoveries. Despite many challenges, genetic research in NEC is expected to increase, providing new insights
into its pathogenesis and bringing the promise of personalized care closer to reality. In this review we provide a
summary of genetic studies in NEC along with defining the challenges and possible future approaches.
1. Introduction technological advances in genotyping platforms, are beginning to yield

promising, if preliminary results. In this review, we (i) summarize
Necrotizing enterocolitis (NEC) is an inflammatory disease of the current genetic studies in NEC, (ii) discuss various challenges and
preterm intestinal tract that afflicts 5–10% of infants weighing < limitations of these studies, and (iii) discuss future approaches for
1500 g at birth [1]. Despite considerable advances in neonatal medi- identifying pathogenic loci for NEC.
cine, outcomes for NEC have remained stagnant and may even be
worsening [2,3]. Many risk factors, such as formula feeds, infection, 2. Candidate gene approach
ischemia, and gut dysbiosis have been implicated in NEC [4]. Clinical
risk factors alone, however, cannot explain the inter-individual varia- The candidate gene approach has been used in the majority of ge-
bility in NEC susceptibility or severity in premature infants, indicating a netic studies of NEC. In this approach, investigators evaluate single
potential role for inherited risk factors in disease pathogenesis [5]. nucleotide polymorphisms (SNPs) of a few, pre-selected genes based on
Epidemiological studies were the first to provide clues regarding a a-priori hypothesis of relevance to NEC. The discussion that follows,
genetic contribution to NEC. NEC risk in the preterm population varies along with Table 1 and Fig. 1, summarizes candidate genes for NEC that
widely based on ethnicity, occurring less frequently in Japan, have been evaluated using this approach, with an emphasis on under-
Switzerland, and Austria while afflicting preterm infants more often in standing the biological relevance of the gene or pathway to NEC de-
North America, UK, and Ireland [4]. Although this variation could be velopment.
due to differences in local neonatal care practices, studies limited to the
USA also demonstrated differences in risk based on race, with African- 2.1. Toll-like receptors (TLRs)
American infants more likely to develop NEC than Caucasians [6,7].
Twin studies by Bhandari et al. [5] also demonstrated that more than TLRs are pattern recognition receptors (PRR) that recognize con-
50% of variance in NEC is accounted for by genetic and shared en- served structural motifs in pathogens and activate innate immune re-
vironmental factors. These studies triggered investigations by several sponses. TLR4, which is important for immunity against Gram-negative
researchers into the genetic basis of NEC risk. Advances in our under- bacteria, has been implicated in NEC by studies showing elevated ex-
standing of the genetic basis of complex diseases, combined with pression in intestinal samples obtained from NEC infants [31].
∗
Corresponding author. Division of Neonatology, Department of Pediatrics, Children's Mercy Kansas City, 2401 Gillham Road, Kansas City, MO, 64108, USA.
E-mail address: vsampath@cmh.edu (V. Sampath).
A. Cuna et al. Seminars in Fetal and Neonatal Medicine 23 (2018) 387–393
Table 1 Table 1 (continued)

Summary of genes assessed using the candidate gene approach.
Gene/pathway Increased risk for Decreased risk No evidence of
Gene/pathway Increased risk for Decreased risk No evidence of NEC for NEC association with
NEC for NEC association with NEC
NEC
HB-EGF [30] HBEGF HBEGF
TLRs [8–10] MD2 (C-1625G) TLR2 (Arg753Gln) (140338940G→T) (140333293G→A)
TLR4 (Asp299Gly) HBEGF
TLR4 Thr399Ile) (140334979C→T)
TLR4 (Asp299Gly) HBEGF
TLR4 (Thr399Ile) (140337403G→C)
TLR5 (Arg392stop) HBEGF
TLR9 (2849 C→T) (140347675G→A)
CD14 (−260 C→T)
IRAK1 (Leu532Ser) NEC, necrotizing enterocolitis.
TIRAP (Ser180Leu)
Genes listed as gene name (single nucleotide polymorphism or amino acid
SIGIRR [11] SIGIRR (p.Y168X)
change).
SIGIRR (p.S80Y) a
NOD Like NOD2 (G908R)a ATG16L1 NOD2 (Gly908Arg) Increased risk for NEC reported when more than two NOD2 variants are
Recptors NOD2 (R702W)a (Thr300Ala) NOD2 (R702W) present.
(NODs)/ NOD2 (1007fs)a NOD2 (1007fs)
Autophagy NOD2 (G908R) Experimental studies in rodents also demonstrated an obligatory role
[8,12–14] NOD2 (rs2066844)
for TLR4 in NEC, further highlighting its importance in disease patho-
NOD1 (37684 T→C)
NLRP3
genesis [32,33]. Preliminary genetic studies in humans, however, have
(247617036G→T) yet to yield strong evidence that genetic variations in TLR pathway
NLRP3 increase susceptibility to NEC. In a cohort of 270 preterm infants,
(247621033G→A) Sampath et al. [9] found no association between NEC and SNPs in TLR
NLRP3
receptors and TLR pathway mediators. Szebani et al. [8] also found no
(14384C→A)
CARD8 association between NEC and SNPs in TLR4 and cluster of differentia-
(26498 T→A) tion 14 (CD14, a key co-receptor of TLR4). In another study, Zhou et al.
MBL [15] MBL2 (−221 G→C) MBL2 (codon [10] investigated myeloid differentiation-2 (MD2, another co-receptor
54 G→A) MBL2
of TLR4) in 42 NEC cases and 83 controls. Overall, there was no dif-
(codon 57G→A)
PAF [10,16] GM2A PAFAH (Ile198Thr)
ference in the frequency of MD2 SNPs between NEC cases and controls;
(g.5220G→A) although analysis based on severity demonstrated a small but sig-
GM2A PAFAH (Ala379Val) nificantly increased risk of surgical NEC with MD2 variants (OR: 1.66;
(g.19168A→C) 95% CI: 1.09–3.05; P = 0.036).
NFKB [9] NFKB1 NFKBIA
(−24519delATTG) (−1004A→G)
Cytokines IL6 (g.4880C→G) IL4RA IL1B (−31 T→C) 2.2. Single immunoglobulin and toll-interleukin 1 receptor (SIGIRR)
[17–22] (g.54150A→G) IL1B (−511C→T)
IL1B (3953 C→T) Using whole exome sequencing (WES), Sampath et al. [11] identi-
IL1B (c.–198 T→C)
fied a novel stop variant (p.Y168X) of SIGIRR in an infant with NEC
IL4RA (c.1902
A→G) totalis. Additional Sanger sequencing of the SIGIRR gene in 37 preterm
IL6 (−174G→C) infants (17 with NEC) demonstrated that rare and novel variants of
IL6 (−596G→A) SIGIRR were enriched in preterm infants with NEC compared to con-
IL8 (−251 T→A) trols, further implicating SIGIRR as a candidate gene for NEC. Func-
IL10 (−1082G→A)
tional studies in intestinal epithelial cells showed that the identified
IL12 (g.19532A→C)
IL18 (−607C→A) variants were functional mutations that led to loss of SIGIRR function,
IL18 (−137G→C) dysregulated TLR activation, and excessive inflammation. Recently, this
PXR (g.39403G→A) group performed invivo studies in SIGIRR–/– mice to demonstrate that
TNF (−238G→A)
loss of SIGIRR in the newborn intestine results in exaggerated intestinal
TNF (8156G→A)
TNF (−308G→A) inflammation and increased NEC susceptibility [34]. Intriguingly,
FUT2 [23,24] FUT2 (428G→A) SIGGIR functions as a “brake” to prevent excessive TLR activation and
Antioxidants HMOX1 (p.D7H) inflammation in the intestines [35]. Thus, although no genes that
[25,26] GCLC (g.–129 C→T) mediate TLR signaling have been strongly associated with NEC, in-
SOD2 (p.A16V)
hibitors of TLR signaling (such as SIGIRR) may be promising NEC
GSTP1 (p.I1047V)
NQO1 (p.P187S) candidate genes.
NFE2L2
(−617C→A) 2.3. Nucleotide binding oligomerization domain-containing protein 2
CYBA (C242T)
(NOD2)
Vascular VEGF (C-2578A) VEGF (T–460C)
homeostasis eNOS (894G→T) VEGF (+405C)
[22,27–29] eNOS (−786 T→C) CPS1 NOD2 is another PRR that recognizes the muramyl dipeptide com-
(p.Thr1406Asn) ponent of bacterial cell wall and interacts with autophagy pathways to
END-1 (5665G→T) target and degrade invading bacteria [36]. Functional polymorphisms
in NOD2, which are strongly implicated in inflammatory bowel disease,
have been hypothesized also to be important in NEC [37]. In a pre-
liminary study, Zouali et al. [12] sequenced the entire NOD2 gene of 10
preterm infants with NEC but found no novel NOD2 variants in NEC
infants. Szebani et al. [8] also found no evidence of association between
SNPs of NOD2 and NEC in their study of 118 preterm infants, 41 of
388
Fig. 1. Risk factors traditionally implicated in necrotizing enterocolitis (NEC) with respect to recognized pathological pathways activated in NEC, and genetic
variants reported to be associated with increased or decreased NEC risk. Red text: increased risk for NEC; green text: protection against NEC. PDA, patent ductus
arteriosus; IUGR, intrauterine growth restriction; IL, interleukin; TLR, Toll-like receptors; eNOS, endothelial nitric oxide synthase; VEGF, vascular endothelial growth
factor; MD2, myeloid differentiation-2; SIGIRR, single immunoglobulin and Toll-interleukin 1 receptor (SIGIRR); NOD2, nucleotide binding oligomerization domain-
containing protein 2; NFKB1, nuclear factor-kappa beta 1; MBL, mannose-binding lectin; ATG, autophagy-related genes; GM2, ganglioside activator; HHBEGF, human
heparin-binding epidermal growth factor; TGFβ, transforming growth factor-β.
whom had NEC. Similar findings of no association were found by found in intestinal tract of preterm infants with NEC, suggesting that
Sampath et al. [14] between NOD1 or NOD2 variants and NEC in a excessive MBL activation may play a role in its pathogenesis [40].
large, multi-center study involving 1015 preterm infants. In an even Prencipe et al. [15] evaluated 107 preterm infants (41 with NEC) for
larger study, Härtel et al. [13] genotyped more than 9000 preterm in- SNPs of MBL and found that a promoter SNP (−221 X/Y) of the MBL
fants of European ancestry and found that whereas no single NOD2 gene was more common in infants with NEC (OR: 4.42; 95% CI:
variant was associated with NEC, presence of two or more loss of 1.11–17.49; P = 0.03). This polymorphism is associated with increased
function NOD2 variants was significantly associated with an increased serum MBL levels, supporting the hypothesis that gain of function
risk for NEC (OR: 4.14; 95% CI: 1.41–12.12; P = 0.009). Thus, whereas variants of MBL is associated with an increased risk for NEC.
initial studies were negative, the largest study seems to suggest that
NOD2 variants confer increased NEC risk in premature infants. 2.6. Platelet activating factor (PAF)
2.4. Autophagy-related 16-like 1 (ATG16L1) PAF is an endogenous phospholipid with potent pro-inflammatory
actions [41]. Infants with NEC have increased plasma levels of PAF, and
ATG16L1 is essential for the proper assembly of lysosomes used in administration of intravenous PAF in animal models induces NEC [42].
the autophagy of invading pathogens. Induction of autophagy is asso- Conversely, PAFAH – an enzyme that degrades PAF – has decreased
ciated with worse injury in experimental NEC, whereas diminished activity and concentration in infants with NEC, and enteral adminis-
autophagy decreases NEC [38]. Sampath et al. [14] evaluated the key tration of recombinant PAFAH in animal models results in a significant
autophagy gene, ATG16L1, and found that a common ATG16L1 variant decrease in NEC [43,44]. Sankararaman et al. [16] evaluated the role of
(Thr300Ala) conferred protection against NEC (OR: 0.4; 95% CI: functional PAFAH polymorphisms in a convenience cohort of 570
0.19–0.81). This variant, which increases sensitization of ATG16L1 to preterm infants (36 with stage 1 or 2 NEC, 534 without NEC) and found
caspase-3-mediated degradation, results in diminished autophagy and no evidence of association between PAFAH SNPs and NEC. In contrast,
is thus consistent with prior studies showing protective role of deceased Zhou et al. [10] evaluated SNPs of GM2 activator protein (GM2A),
autophagy against NEC. An independent replication cohort (260 in- which is another PAF antagonist, in a Chinese cohort of 125 infants (42
fants, 23 with NEC) also showed a trend towards decreased NEC among with NEC) and found a positive association between NEC and two
infants who were homozygous for the loss of function variant allele, variants of GM2A – rs1048719 (OR: 1.86; 95% CI: 1.04–3.33) and
providing further evidence for the importance of this gene in NEC [14]. rs2075783 (OR: 3.23; 95% CI: 1.69–6.21). Inclusion of term infants in
Further studies are needed to evaluate the role of this and other au- this study, however, makes it difficult to determine whether the dif-
tophagy-related genes in NEC. ferences detected were due to prematurity or NEC.
2.5. Mannose-binding lectin (MBL) 2.7. Nuclear factor-kappa beta (NFKB)
MBL is a circulating PRR that recognizes unique carbohydrate NFKB is a major transcription factor that regulates the expression of
components of invading pathogens and induces their phagocytic several pro-inflammatory genes [45]. NFKB levels are increased in the
clearance by the complement system [39]. Elevated MBL levels are intestines following NEC induction in rats, whereas inhibiting NFKB
389
protects against excessive inflammation [46,47]. Sampath et al. [9] state for VEGF-2578 mutant allele, which predisposes to low VEGF
found that the NFKB1 (g.-24519delATTG) variant was present in all 15 production, is associated with an increased risk for NEC (OR: 2.77; 95%
infants with NEC compared to only 166 of 256 without NEC (100% vs CI: 1.00–7.65; P = 0.049). This is consistent with studies in mice de-
65%; OR: ∞; 95% CI: 2.1–∞; P = 0.003), suggesting that this variant monstrating that decreased VEGF signaling increases susceptibility to
may be necessary, but not sufficient, for developing NEC. On the other NEC, whereas increased VEGF levels through subcutaneous adminis-
hand, the NFKBIA (g.1004A→G) variant was less common in infants tration decreases the extent of mucosal injury in NEC [50,51].
with NEC compared to infants without NEC (13.3% vs 49%; OR: 0.16;
95% CI: 0.04–0.60; P = 0.007), suggesting that this variant may confer 2.12. Arginine and nitric oxide
protection against NEC.
Arginine is an important substrate for nitric oxide, which is a potent
2.8. Pro-inflammatory cytokines vasodilator that regulates mucosal blood flow and helps maintain mu-
cosal integrity. Low arginine levels have been implicated in NEC based
As excessive inflammation is a hallmark of NEC, many investigators on studies demonstrating hypoarginemia in preterm infants with NEC,
have examined the relationship of several pro-inflammatory cytokine and clinical trials showing decreased NEC in infants following arginine
gene polymorphisms and NEC, with varying results. Treszl et al. supplementation [52,53]. Moonen et al. [27] evaluated C→A nucleo-
[17,18], Heninger et al. [19], Henderson et al. [20], and Szpecht et al. tide transversion (T1405 N) of carbamoyl-phosphate synthase (CPS),
[22] have examined functional variants of several pro-inflammatory which is functionally correlated with low plasma arginine levels, in a
cytokines (including tumor necrosis factor-α, interleukin (IL)-1β, IL-4 small study of 51 preterm infants (17 with NEC). They found that the
receptor α-chain, IL-6, IL-18, and IL-10) and found no evidence of as- CC genotype was associated with increased NEC (unadjusted OR: 3.43;
sociation of these genes with NEC. In contrast, Franklin et al. [21] re- 95% CI: 1.01–11.49), although this association was not significant once
ported a positive association between IL-6 (rs1800795) variant and adjusted for gestational age and birth weight. As a follow-up to this
NEC in Caucasian neonates; and Tian et al. [48] found a positive as- study, Moonen et al. [28] prospectively recruited 477 preterm infants
sociation between IL17F (rs763780) variant and NEC in a Chinese co- from four different centers to investigate the same CPS1 polymorphism
hort. Interestingly, the studies showing no association between pro- in NEC. They again found no significant association between CC gen-
inflammatory cytokines and NEC was comprised of neonates of Eur- otype and NEC. Instead, a significant negative association was found
opean descent, suggesting that ethnicity may play an important role in between the minor A-allele and the combined outcome of NEC or death,
whether certain genetic variants contribute to NEC. suggesting that the A-allele variant may confer protection against NEC.
In another study, Szpecht et al. [22] more directly evaluated nitric
2.9. Fucosyltransferase 2 (FUT2) oxide pathways by measuring polymorphisms of endothelial nitric
oxide synthase (eNOS) in 100 preterm infants, 22 of whom had NEC.
FUT2 mediates the inclusion of fucose sugar units to glycoproteins They found that eNOS 894G→T polymorphism was more common in
and glycolipids. Polymorphisms in FUT2 gene, which result in differing infants with NEC (OR: 20; 95% CI: 3.71–208.7; P = 0.0004), and that
phenotypes of secreted fucosylated glycans on mucosal surfaces, have the eNOS 786 T→C was increased in infants with surgical NEC (OR:
important implications in host–microbiome interactions. Morrow et al. 4.88; 95% CI 1.33–21.99; P = 0.013). Both variants have been asso-
[23] investigated non-secretor status (AA) of the FUT2 polymorphism ciated in other studies with impaired eNOS enzymatic activity, pro-
(428G→A) and found no evidence of association between this variant viding biological plausibility that these variants are functional and
and NEC (OR: 1.6; 95% CI: 0.7–3.8) or surgical NEC (OR: 2.3; 95% CI: potentially significant in NEC [54].
0.6–7.4). However, when salivary secretion of H-antigen was used to
determine non-secretor status, a positive association with NEC was 2.13. Heparin-binding epidermal growth factor-like growth factor (HB-
found. Another study by Demmert et al. [24] investigated the same EGF)
FUT2 polymorphism (248G→A) in a large prospective cohort of 2406
VLBW infants and found no association of this polymorphism with NEC. HB-EGF has been shown in animal studies to promote repair of in-
These differential results can be possibly explained by varied influence testinal epithelial cells resulting from various forms in injury in vitro,
of the FUT2 genotype on the secretor phenotype, which has been shown including in NEC [55]. The protective role of HB-EGF is thought to be
to be influenced by epigenetic changes, and not always concordant with mediated by increased intestinal blood flow, decreased apoptosis, and
genotypes. increased enterocyte migration and proliferation [56]. Ma et al. [30]
measured polymorphisms of HB-EGF in 30 NEC patients and 80 con-
2.10. Genes regulating the antioxidant response trols, finding that rs4912711 SNP was associated with increased NEC.
They also measured plasma levels of HB-EGF and found it to be de-
Antioxidants are important in alleviating injury from gut ische- creased in NEC infants with the rs4912711 SNP, providing additional
mia–reperfusion, and genetic variants that limit antioxidant defenses evidence that the identified SNP is potentially important.
can hypothetically increase risk of NEC in preterm infants [49]. Sam-
path et al. [25] investigated loss of function SNPs in six antioxidant 3. Genome-wide approach
genes belonging to the NF-E2-related factor 2-dependent antioxidant
response elements (Nrf2-ARE) pathway and found no evidence of as- In the genome-wide approach, millions of SNPs are evaluated in an
sociation between the tested SNPs and NEC. Another study by Huizing unbiased manner, allowing for discovery of previously unknown asso-
et al. [26] looked at the C242T polymorphism of cytochrome B-245 ciations between genes and disease. Jilling et al. [57] performed the
alpha chain (a gene related to NADPH oxidase family), also finding no first genome-wide association study (GWAS) in NEC in 751 extremely
association of this variant with NEC in a cohort of 451 preterm infants. preterm infants (30 with surgical NEC) and identified 35 SNPs sig-
nificantly associated with NEC. A cluster of SNPs in chromosome 8
2.11. Vascular endothelial growth factor (VEGF) (8q23.3) had the strongest association for NEC (OR: 4.72 (95% CI:
2.51–8.88). Other clusters of SNPs identified in chromosomes 14 and 11
Studies demonstrate that decreased intestinal expression of VEGF is also demonstrated significant association. A validation cohort did not
seen in human NEC compared to controls, suggesting that dysregulated have sufficient power to replicate these findings, except for a single SNP
VEGF signaling contributes to NEC pathogenesis. Banyasz et al. [29] found in chromosome 8. Additional studies are required to confirm
evaluated VEGF polymorphisms in NEC and discovered that a carrier functional significance of the identified SNPs. Further fine mapping or
390
Table 2
Characteristics of included studies.
Study Country No. of cases/controls Definition of NEC cases Functional studies Validation cohort
Trezl et al. [18] Hungary 46/90 Included Bell stage I No No

Heninger et al. [19] Hungary 46/90 Included Bell stage I No No
Treszl et al. [17] Hungary 46/90 Included Bell stage I No No
Zouali et al. [12] France 10/103 Unclear No No
Szebani et al. [8] Hungary 41/77 Included Bell stage I No No
Banyasz et al. [29] Hungary 49/79 Included Bell stage I No No
Moonen et al. [27] Netherlands 17/34 ≥ Bell stage II No No
Henderson et al. [20] Australia 50/50 Included Bell stage I No No
Sampath et al. [9] USA 15/255 Included Bell stage I No No
Morrow et al. [23] USA 30/410 ≥ Bell stage II Yes No
Prencipe et al. [15] Italy 41/66 ≥ Bell stage II Yes No
Sankararaman et al. [16] USA 36/534 Included Bell stage I No No
Zhou et al. [10] China 42/83 ≥ Bell stage II No No
Franklin et al. [21] USA 66/118 ≥ Bell stage 2 No No
Sampath et al. [11] USA 17/20 ≥ Bell stage II Yes No
Sampath et al. [25] USA 52/585 ≥ Bell stage II No No
Demmert et al. [24] Germany 108/2406 Included SIP and surgical NEC No No
Hartel et al. [13] Germany 262/6680 Included SIP and ≥ Bell stage II No No
Moonen et al. [28] Netherlands 36/441 ≥ Bell stage II No Yes
Tian et al. [48] China 102/120 ≥ Bell stage II No No
Huizing et al. [26] Netherlands 28/392 ≥ Bell stage II No No
Sampath et al. [14] USA 86/929 Included Bell stage I No Yes
Ma et al. [30] China 30/80 ≥ Bell stage II Yes No
Szpecht et al. [22] Poland 22/78 Included Bell stage I No No
Jilling et al. [57] USA 30/721 Surgical NEC No Yes
sequencing studies may also be necessary to identify pathogenic genes sequencing, where the significance of identified variants to NEC pa-
or variants in identified clusters. thogenesis needs to be determined.
4. Challenges and limitations 5. Future considerations
Several challenges and limitations can confound the results of ge- Advances in genetic research have allowed identification of many
netic association studies, especially for complex diseases such as NEC promising genes associated with increased or decreased risk of NEC.
[58]. The following discussion summarizes these challenges as well as Most of these studies, however, have been limited by small sample sizes
important considerations for performing robust genetic studies in NEC. as well as lack of validation and functional studies. Efforts to validate
Table 2 characterizes the different genetic studies included in this re- these candidate genes, using independent replication cohorts and
view based on these criteria. A precise phenotypic definition of NEC functional studies, are necessary. Although challenging, such studies
cases and controls is essential. Only confirmed cases of NEC in preterm are needed to establish the importance of these genes in conferring
infants (i.e. Bell's stage ≥ II) should be included. Inclusion of term in- increased NEC risk.
fants, spontaneous intestinal perforations (SIP), or unconfirmed NEC Future studies should also utilize unbiased genome-wide ap-
(Bell stage I) as NEC cases can limit the study's power. Further confining proaches, such as GWAS, WES, or whole genome sequencing (WGS), to
to infants who developed surgical NEC may confer additional precision discover new genes and pathways. Linkage-based approaches used in
to the phenotype. Ethnic composition of the cohort should also be ac- GWAS typically identify common variants (present in > 1% of the po-
curately described. Variant allele frequencies can vary significantly by pulation) in intronic regions. Because of this, additional fine mapping/
ethnicity, and false associations may arise because of population ad- sequencing as well as functional studies are typically needed to identify
mixture. In cohorts with mixed ethnicity, additional analysis of variants pathogenic genes and variants. In contrast, WES directly identifies
stratified by race should be completed. variants in the coding or exonic regions that are more likely to have a
An adequate sample size is also key. Ideally, candidate genome functional impact on the gene product. WES is thus increasingly
approaches should include hundreds of infants, while genome-wide emerging as the primary modality to examine the genetic basis of
approaches often need thousands of infants to attain sufficient power. complex diseases [59–61]. WGS probes both coding and non-coding
Such numbers are difficult to achieve for any disease, and this is regions of the genome [62]. However, WGS encumbers very large
especially difficult in NEC which variably afflicts about 5–10% of in- samples sizes; and functional evaluation of regulatory variants found in
fants born preterm. Compounding this challenge is the need to correct non-coding regions (i.e. introns, promoters, etc.) is more challenging
for multiple testing across the number of SNPs tested. Due to the [63]. Thus, combining GWAS with fine mapping, or alternatively WES
challenges with sample size, only a handful of studies have been able to directly, are suitable approaches to identify pathogenic loci for NEC.
adjust for multiple testing. Contribution of other “-omic” fields, which can be helpful in iden-
Separate validation cohorts are also important. In general, genetic tifying genetic signatures underlying NEC, should also be pursued. For
association studies should be treated as preliminary, hypothesis-gen- example, Chan et al. [64] performed transcriptome (mRNA) sequencing
erating studies until further validation has been provided by subsequent of intestinal tissues from infants with surgical NEC compared to con-
replication cohorts. Due to the challenge of obtaining adequate sample trols and identified several dysregulated genes and pathways in NEC.
size for discovery cohorts, few genetic studies in NEC have been re- Using a similar strategy, Ng et al. [65] probed microRNAs (miR), a
plicated in validation cohorts. Lastly, functional studies that explain the subset of small RNAs which are not protein-coding but which regulate
mechanisms underlying the association of the gene with the disease are other mRNAs, to identify regulatory pathways of interest in NEC. In-
also important. Functional studies are particularly important for terestingly, both of these studies identified the TLRs as a key pathway
genome-wide approaches such as GWAS and exome/whole genome involved in NEC causation. This discovery, which further highlights the
391
importance of TLR in NEC, also provides evidence supporting the use- References
fulness of this approach to identify potential genetic pathways in NEC.
Finally, functional genomics approaches that combine genomics [1] Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011;364:255–64.
with transcriptome and proteomics approaches may provide even [2] Patel RM, Kandefer S, Walsh MC, et al. Causes and timing of death in extremely
premature infants from 2000 through 2011. N Engl J Med 2015;372:331–40.
deeper insights into understanding how pathogenic genes direct mole- [3] Blakely ML, Lally KP, McDonald S, et al. Postoperative outcomes of extremely low
cular phenotypes that program NEC vulnerability [66,67,68]. Major birth-weight infants with necrotizing enterocolitis or isolated intestinal perforation:
advances in bioinformatics and network analysis provide platforms to a prospective cohort study by the NICHD Neonatal Research Network. Ann Surg
2005;241:984–9. discussion 989–994.
interrogate and define the key genetic pathways as well as molecular [4] Zani A, Pierro A. Necrotizing enterocolitis: controversies and challenges. F1000 Res
signatures that prime NEC vulnerability. 2015;4.
[5] Bhandari V, Bizzarro MJ, Shetty A, et al. Familial and genetic susceptibility to major
neonatal morbidities in preterm twins. Pediatrics 2006;117:1901–6.
6. Conclusion [6] Llanos AR, Moss ME, Pinzon MC, Dye T, Sinkin RA, Kendig JW. Epidemiology of
neonatal necrotising enterocolitis: a population-based study. Paediatr Perinat
Epidemiol 2002;16:342–9.
It is expected that genetic studies in NEC will continue to increase as [7] Uauy RD, Fanaroff AA, Korones SB, Phillips EA, Phillips JB, Wright LL. Necrotizing
technological advances allow for more robust and less expensive ex- enterocolitis in very low birth weight infants: biogeographic and clinical correlates.
National Institute of Child Health and Human Development Neonatal Research
ploration of the genome. Although potential gene variants have been
Network. J Pediatr 1991;119:630–8.
identified, these discoveries have to be interpreted with caution until [8] Szebeni B, Szekeres R, Rusai K, et al. Genetic polymorphisms of CD14, toll-like
further validation studies are completed. Future directions include use receptor 4, and caspase-recruitment domain 15 are not associated with necrotizing
of whole exome/genome sequencing to uncover novel and rare genes/ enterocolitis in very low birth weight infants. J Pediatr Gastroenterol Nutr
2006;42:27–31.
variants in NEC. Complementary functional genomic approaches may [9] Sampath V, Le M, Lane L, et al. The NFKB1 (g.-24519delATTG) variant is associated
allow NEC prediction in premature infants as well as defining pathways with necrotizing enterocolitis (NEC) in premature infants. J Surg Res
that may be targeted to prevent NEC. Although these exciting ap- 2011;169:e51–7.
[10] Zhou W, Yuan W, Huang L, Wang P, Rong X, Tang J. Association of neonatal ne-
proaches hold significant promise for defining the molecular patho- crotizing enterocolitis with myeloid differentiation-2 and GM2 activator protein
genesis of NEC and informing preventive approaches, they require genetic polymorphisms. Mol Med Rep 2015;12:974–80.
collaboration between clinicians, researchers, and bioinformaticians for [11] Sampath V, Menden H, Helbling D, et al. SIGIRR genetic variants in premature
infants with necrotizing enterocolitis. Pediatrics 2015;135:e1530–4.
their success. [12] Zouali H, Bonnard A, De Lagausie DL, et al. CARD15/NOD2 is not a predisposing
factor for necrotizing enterocolitis. Dig Dis Sci 2005;50:1684–7.
[13] Hartel C, Hartz A, Pagel J, et al. NOD2 loss-of-function mutations and risks of ne-
6.1. Practice points crotizing enterocolitis or focal intestinal perforation in very low-birth-weight in-
fants. Inflamm Bowel Dis 2016;22:249–56.
• Presence of clinical risk factors does not predict NEC occurrence or [14] Sampath V, Bhandari V, Berger J, et al. A functional ATG16L1 (T300A) variant is
associated with necrotizing enterocolitis in premature infants. Pediatr Res
severity in premature infants, suggesting a role for genetic risk 2017;81:582–8.
factors. [15] Prencipe G, Azzari C, Moriondo M, et al. Association between mannose-binding
• A “two-hit” model involving interactions between genetic and lectin gene polymorphisms and necrotizing enterocolitis in preterm infants. J
Pediatr Gastroenterol Nutr 2012;55:160–5.
clinical risk factors may be required for NEC causation. [16] Sankararaman S, Yanamandra K, Napper D, Caldito G, Dhanireddy R. The pre-
• Identification of genetic variants that prime NEC risk may lead to valence of platelet activating factor acetylhydrolase single nucleotide polymorph-
isms in relationship to necrotizing enterocolitis in Northwest Louisiana infants.
the development of screening-based approaches to identify suscep-
SpringerPlus 2013;2:294.
tible infants.
•
[17] Treszl A, Heninger E, Kalman A, Schuler A, Tulassay T, Vasarhelyi B. Lower pre-
New knowledge of genetic pathways regulating NEC risk may in- valence of IL-4 receptor alpha-chain gene G variant in very-low-birth-weight infants
form development of novel mechanism-based therapies to prevent with necrotizing enterocolitis. J Pediatr Surg 2003;38:1374–8.
[18] Treszl A, Kocsis I, Szathmari M, Schuler A, Tulassay T, Vasarhelyi B. Genetic var-
disease. iants of the tumour necrosis factor-alpha promoter gene do not influence the de-
velopment of necrotizing enterocolitis. Acta Paediatr 2001;90:1182–5.
[19] Heninger E, Treszl A, Kocsis I, Derfalvi B, Tulassay T, Vasarhelyi B. Genetic variants
6.2. Research directions of the interleukin-18 promoter region (-607) influence the course of necrotising
enterocolitis in very low birth weight neonates. Eur J Pediatr 2002;161:410–1.
• Accelerating genome-wide sequencing approaches to uncover novel [20] Henderson G, Craig S, Baier RJ, Helps N, Brocklehurst P, McGuire W. Cytokine gene
polymorphisms in preterm infants with necrotising enterocolitis: genetic association
candidate genes and pathways for NEC susceptibility. study. Arch Dis Child Fetal Neonatal Ed 2009;94:F124–8.
• Emphasis of precise phenotype definition, adequate validation co- [21] Franklin AL, Said M, Cappiello CD, et al. Are immune modulating single nucleotide
polymorphisms associated with necrotizing enterocolitis? Sci Rep 2015;5:18369.
horts, and functional analysis.
• Combining genetic approaches with transcriptome, proteome and
[22] Szpecht D, Neumann-Klimasinska N, Blaszczynski M, et al. Candidate gene analysis
in pathogenesis of surgically and non-surgically treated necrotizing enterocolitis in
microbiome analysis to comprehensively evaluate the molecular preterm infants. Mol Cell Biochem 2018;439:53–63.
basis of the NEC phenotype. [23] Morrow AL, Meinzen-Derr J, Huang P, et al. Fucosyltransferase 2 non-secretor and
low secretor status predicts severe outcomes in premature infants. J Pediatr
2011;158:745–51.
[24] Demmert M, Schaper A, Pagel J, et al. FUT 2 polymorphism and outcome in very-
Conflicts of interest statement low-birth-weight infants. Pediatr Res 2015;77:586–90.
[25] Sampath V, Garland JS, Helbling D, et al. Antioxidant response genes sequence
None declared. variants and BPD susceptibility in VLBW infants. Pediatr Res 2015;77:477–83.
[26] Huizing MJ, Cavallaro GH, Moonen RM, et al. Is the C242T polymorphism of the
CYBA gene linked with oxidative stress-associated complications of prematurity?
Antioxidants Redox Signal 2017;27:1432–8.
Funding sources [27] Moonen RM, Paulussen AD, Souren NY, Kessels AG, Rubio-Gozalbo ME, Villamor E.
Carbamoyl phosphate synthetase polymorphisms as a risk factor for necrotizing
VS's effort was partly supported by 1 R01 DK117296-01A1 (VS). enterocolitis. Pediatr Res 2007;62:188–90.
[28] Moonen RM, Cavallaro G, Huizing MJ, Gonzalez-Luis GE, Mosca F, Villamor E.
Association between the p.Thr1406Asn polymorphism of the carbamoyl-phosphate
synthetase 1 gene and necrotizing enterocolitis: a prospective multicenter study. Sci
Acknowledgements
Rep 2016;6:36999.
[29] Banyasz I, Bokodi G, Vasarhelyi B, et al. Genetic polymorphisms for vascular en-
1. Institutional start-up funds from Children's Mercy Kansas City dothelial growth factor in perinatal complications. Eur Cytokine Netw
(VS); 2. Ken and Eva Smith Clinical Scholar Award (AC), Kansas City, 2006;17:266–70.
[30] Ma F, Li S, Hao H, et al. Association of heparin-binding EGF-like growth factor
MO.
392
polymorphisms with necrotizing enterocolitis in preterm infants. J Pediatr Perinatol 2015;32:905–9.

Gastroenterol Nutr 2018 Apr;66(4):e99–102. [50] Yan X, Managlia E, Liu SX, et al. Lack of VEGFR2 signaling causes maldevelopment
[31] Leaphart CL, Cavallo J, Gribar SC, et al. A critical role for TLR4 in the pathogenesis of the intestinal microvasculature and facilitates necrotizing enterocolitis in neo-
of necrotizing enterocolitis by modulating intestinal injury and repair. J Immunol natal mice. Am J Physiol Gastrointest Liver Physiol 2016. ajpgi 00273 02015.
2007;179:4808–20. [51] Karatepe HO, Kilincaslan H, Berber M, et al. The effect of vascular endothelial
[32] Jilling T, Simon D, Lu J, et al. The roles of bacteria and TLR4 in rat and murine growth factor overexpression in experimental necrotizing enterocolitis. Pediatr Surg
models of necrotizing enterocolitis. J Immunol 2006;177:3273–82. Int 2014;30:327–32.
[33] Lu J, Jilling T, Li D, Caplan MS. Polyunsaturated fatty acid supplementation alters [52] Zamora SA, Amin HJ, McMillan DD, et al. Plasma L-arginine concentrations in
proinflammatory gene expression and reduces the incidence of necrotizing en- premature infants with necrotizing enterocolitis. J Pediatr 1997;131:226–32.
terocolitis in a neonatal rat model. Pediatr Res 2007;61:427–32. [53] Polycarpou E, Zachaki S, Tsolia M, et al. Enteral L-arginine supplementation for
[34] Fawley J, Cuna A, Menden HL, et al. Single-Immunoglobulin Interleukin-1-Related prevention of necrotizing enterocolitis in very low birth weight neonates: a double-
Receptor regulates vulnerability to TLR4-mediated necrotizing enterocolitis in a blind randomized pilot study of efficacy and safety. JPEN. JPEN (J Parenter Enteral
mouse model. Pediatr Res 2018;83:164–74. Nutr) 2013;37:617–22.
[35] Wald D, Qin J, Zhao Z, et al. SIGIRR, a negative regulator of Toll-like receptor- [54] Veldman BA, Spiering W, Doevendans PA, et al. The Glu298Asp polymorphism of
interleukin 1 receptor signaling. Nat Immunol 2003;4:920–7. the NOS 3 gene as a determinant of the baseline production of nitric oxide. J
[36] Chen G, Shaw MH, Kim YG, Nunez G. NOD-like receptors: role in innate immunity Hypertens 2002;20:2023–7.
and inflammatory disease. Annu Rev Pathol 2009;4:365–98. [55] Radulescu A, Zhang HY, Yu X, et al. Heparin-binding epidermal growth factor-like
[37] Cho JH, Abraham C. Inflammatory bowel disease genetics: Nod2. Annu Rev Med growth factor overexpression in transgenic mice increases resistance to necrotizing
2007;58:401–16. enterocolitis. J Pediatr Surg 2010;45:1933–9.
[38] Neal MD, Sodhi CP, Dyer M, et al. A critical role for TLR4 induction of autophagy in [56] Feng J, Besner GE. Heparin-binding epidermal growth factor-like growth factor
the regulation of enterocyte migration and the pathogenesis of necrotizing en- promotes enterocyte migration and proliferation in neonatal rats with necrotizing
terocolitis. J Immunol 2013;190:3541–51. enterocolitis. J Pediatr Surg 2007;42:214–20.
[39] Gadjeva M, Thiel S, Jensenius JC. The mannan-binding-lectin pathway of the innate [57] Jilling T, Ambalavanan N, Cotten CM, et al. Surgical necrotizing enterocolitis in
immune response. Curr Opin Immunol 2001;13:74–8. extremely premature neonates is associated with genetic variations in an intergenic
[40] Schlapbach LJ, Aebi C, Fisch U, et al. Higher cord blood levels of mannose-binding region of chromosome eight. Pediatr Res 2018 May;83(5):943–53.
lectin-associated serine protease-2 in infants with necrotising enterocolitis. Pediatr [58] Marian AJ. Molecular genetic studies of complex phenotypes. Translational re-
Res 2008;64:562–6. search. J Lab Clin Med 2012;159:64–79.
[41] Caplan MS, Sun XM, Hsueh W. Hypoxia causes ischemic bowel necrosis in rats: the [59] Bamshad MJ, Ng SB, Bigham AW, et al. Exome sequencing as a tool for Mendelian
role of platelet-activating factor (PAF-acether). Gastroenterology 1990;99:979–86. disease gene discovery. Nat Rev Genet 2011;12:745–55.
[42] Caplan MS, Sun XM, Hseuh W, Hageman JR. Role of platelet activating factor and [60] Tennessen JA, Bigham AW, O'Connor TD, et al. Evolution and functional impact of
tumor necrosis factor-alpha in neonatal necrotizing enterocolitis. J Pediatr rare coding variation from deep sequencing of human exomes. Science
1990;116:960–4. 2012;337:64–9.
[43] Caplan MS, Hedlund E, Adler L, Lickerman M, Hsueh W. The platelet-activating [61] Kiezun A, Garimella K, Do R, et al. Exome sequencing and the genetic basis of
factor receptor antagonist WEB 2170 prevents neonatal necrotizing enterocolitis in complex traits. Nat Genet 2012;44:623–30.
rats. J Pediatr Gastroenterol Nutr 1997;24:296–301. [62] Bick D, Dimmock D. Whole exome and whole genome sequencing. Curr Opin
[44] Caplan MS, Lickerman M, Adler L, Dietsch GN, Yu A. The role of recombinant Pediatr 2011;23:594–600.
platelet-activating factor acetylhydrolase in a neonatal rat model of necrotizing [63] Panoutsopoulou K, Tachmazidou I, Zeggini E. In search of low-frequency and rare
enterocolitis. Pediatr Res 1997;42:779–83. variants affecting complex traits. Hum Mol Genet 2013;22:R16–21.
[45] Lawrence T. The nuclear factor NF-kappaB pathway in inflammation. Cold Spring [64] Chan KY, Leung KT, Tam YH, et al. Genome-wide expression profiles of necrotizing
Harbor Perspect Biol 2009;1:a001651. enterocolitis versus spontaneous intestinal perforation in human intestinal tissues:
[46] Chung DH, Ethridge RT, Kim S, et al. Molecular mechanisms contributing to ne- dysregulation of functional pathways. Ann Surg 2014;260:1128–37.
crotizing enterocolitis. Ann Surg 2001;233:835–42. [65] Ng PC, Chan KY, Leung KT, et al. Comparative MiRNA expressional profiles and
[47] De Plaen IG, Liu SX, Tian R, et al. Inhibition of nuclear factor-kappaB ameliorates molecular networks in human small bowel tissues of necrotizing enterocolitis and
bowel injury and prolongs survival in a neonatal rat model of necrotizing en- spontaneous intestinal perforation. PLoS One 2015;10:e0135737.
terocolitis. Pediatr Res 2007;61:716–21. [66] Pastinen T. Genome-wide allele-specific analysis: insights into regulatory variation.
[48] Tian J, Liu Y, Jiang Y, et al. Association of single nucleotide polymorphisms of Nat Rev Genet 2010;11:533–8.
IL23R and IL17 with necrotizing enterocolitis in premature infants. Mol Cell [67] Albert FW, Kruglyak L. The role of regulatory variation in complex traits and dis-
Biochem 2017;430:201–9. ease. Nat Rev Genet 2015;16:197–212.
[49] Marseglia L, D'Angelo G, Manti S, et al. Oxidative stress-mediated damage in [68] Stunnenberg HG, Hubner NC. Genomics meets proteomics: identifying the culprits
Newborns with necrotizing enterocolitis: a possible role of melatonin. Am J in disease. Hum Genet 2014;133:689–700.
393

Innate and adaptive immunity in necrotizing enterocolitis T

a a b,∗,1
Madison A. Mara , Misty Good , Joern-Hendrik Weitkamp
a
Department of Pediatrics, Division of Newborn Medicine, St. Louis Children's Hospital, Washington University School of Medicine, St. Louis, MO, USA
b
Department of Pediatrics, Division of Neonatology, Monroe Carell Jr. Children's Hospital at Vanderbilt, Vanderbilt University Medical Center, Nashville, TN, USA
Keywords: Necrotizing enterocolitis (NEC) is the most frequent and devastating gastrointestinal disease of premature in-
Necrotizing enterocolitis fants. Although the precise mechanisms are not fully understood, NEC is thought to develop following a com-
Innate immunity bination of prematurity, formula feeding, and adverse microbial colonization. Within the last decade, studies
Adaptive immunity increasingly support an important role of a heightened mucosal immune response initiating a pro-inflammatory
Intestinal epithelium
signaling cascade, which can lead to the disruption of the intestinal epithelium and translocation of pathogenic
Toll-like receptors
Prematurity
species. In this review, we first describe the cellular composition of the intestinal epithelium and its critical role
T lymphocytes in maintaining epithelial integrity. We then discuss cell signaling during NEC, specifically, toll-like receptors and
nucleotide oligomerization domain-like receptors. We further review cytokines and cellular components that
characterize the innate and adaptive immune systems and how they interact to support or modulate NEC de-
velopment.
1. Introduction protection from micro-organisms. These cell types are joined by tight
junctions and form a crypt-villus structure characteristic of the small
The onset of necrotizing enterocolitis (NEC) is thought to be affected intestine [3]. The villus is covered with enterocytes, goblet, en-
by three primary factors: prematurity, formula feeding, and unbalanced teroendocrine, and tuft cells, whereas the crypts house Paneth cells,
microbial colonization of the intestine [1,2]. However, a prevalent transit amplifying cells, and the progenitor stem cell that gives rise to
unifying hypothesis has proposed that the preterm gut environment is all the intestinal cell types [4]. Together, these cells make up the in-
highly sensitive to postnatal colonization with potentially pathogenic testinal epithelium and act as the first major barrier of the intestinal
bacteria, which elicits inappropriate immune responses supporting NEC innate system.
development [2]. Accordingly, abundant studies indicate that both in-
nate and adaptive immune systems contribute to NEC pathology. By 2.1.1. Enterocytes
reviewing the functions of multiple components of the innate and As absorption is a predominant role of the small intestine, en-
adaptive immune systems in the neonatal intestine, we aim to provide a terocytes (also known as intestinal epithelial cells, IECs) are responsible
comprehensive understanding of the role of the immune system in NEC for uptake of water, nutrients, and vitamins [5]. Enterocytes make up
development. approximately 80% of the intestinal epithelium, are renewed every
three to five days, and are connected by tight junctions, which together
2. Innate immune system maintain intestinal integrity [4,6]. During digestion, enterocytes are
exposed to antigens from food, native microflora, and foreign micro-
2.1. Physical barrier of the intestine organisms [6]. Therefore, enterocytes can sample and identify antigens
passing through the intestinal lumen and relay the signal to the un-
The intestinal epithelium is composed of a complex network with derlying intraepithelial lymphocytes via pattern-recognition receptors
approximately seven different cell types working together to balance (PRRs) [5,7,8]. PRRs expressed by IECs, such as toll-like receptors
the multiple functions of the small intestine, including nutrient ab- (TLRs) and nucleotide-binding oligomerization domains (NODs), re-
sorption, antigen recognition, maintenance of mucosal integrity, and cognize antigens on pathogenic bacteria and elicit an immune response
∗
Corresponding author. Department of Pediatrics, Division of Neonatology, Monroe Carell Jr. Children's Hospital at Vanderbilt, Vanderbilt University Medical
Center, Nashville, TN, 37232, USA.
E-mail address: hendrik.weitkamp@vanderbilt.edu (J.-H. Weitkamp).
1
Doctor's Office Tower, 2200 Children's Way, Nashville, TN 37232, USA.
M.A. Mara et al. Seminars in Fetal and Neonatal Medicine 23 (2018) 394–399
against infection [5,7]. Studies demonstrate that PRRs have a direct determine the risk of developing lethal gastrointestinal tract diseases,
linkage to NEC, which are discussed later in this review. In addition to such as NEC and focal intestinal perforation [26]. Moreover, an un-
PRRs, enterocytes are capable of presenting major histocompatibility conventional mouse model has been described, in which Paneth cells
class (MHC) I and II molecules and non-classical MHC molecules such are depleted with dithizone and the intestine is exposed to Klebsiella
as MHC class I polypeptide-related sequence A or B, both of which can pneumoniae [27–29]. This model yields a phenotype similar to human
signal directly with lymphocytes to initiate an immune response NEC, yet, due to intentional damage of Paneth cells, conclusions re-
[6,7,9]. Taken together, enterocytes play an important role in overall garding AMPs cannot be made [29]. Taken together, further research is
function of the small intestine. needed to elucidate the role of Paneth cells and/or Paneth cell-derived
AMPs in the prevention or the pathogenesis of NEC.
2.1.2. Goblet cells
Goblet cells are an important cell type that help maintain the gut 2.2. Pattern recognition receptors
barrier and are responsible for the production of the mucus layer be-
tween the epithelium and lumen. They make up approximately 4% of Pattern recognition receptors are conserved receptors that have
the epithelium and reside along the crypts up to the villus tip, allowing evolved to sense the presence of pathogenic and endogenous molecules
for mucus secretion [4]. The mucus secreted by goblet cells is composed released during infection and injury. There are currently four classes of
of glycoproteins known as mucins, which are regulated by the mucin PRR families [30], yet for the purpose of this review we focus on the
(MUC) gene. Mucins are responsible for facilitating interactions be- two classes that have been linked to NEC: TLRs and nucleotide oligo-
tween host epithelium and commensal or pathogenic micro-organisms merization domain-like receptors (NLRs).
[9,10]. MUC2 is the primary mucin and important for goblet cell
morphology, as demonstrated by MUC2-deficient mice that fail to de- 2.2.1. Toll-like receptors
velop distinguishable goblet cells [11,12]. Goblet cells have also been TLRs are well-characterized transmembrane receptors that pri-
implicated in immune tolerance to commensal bacteria and are capable marily function to initiate immune responses against bacteria, viruses,
of passaging antigens from the lumen to CD103+ dendritic cells, thus and other pathogenic micro-organisms [30,31]. Presently, there have
promoting intestinal immune homeostasis [13]. been 10 TLRs identified in humans and 13 in mice, with each TLR
Goblet cell differentiation is promoted by kruppel-like factor 4 sensing for a distinct pathogenic or endogenous ligand [30,31]. Once
(Klf4) and E47-like factor 3 (Elf3), but inhibited by the Notch signaling the ligand binds to its associated TLR, the receptor dimerizes and re-
pathway [4,14]. Sodhi et al. demonstrated that toll-like receptor 4 cruits the myeloid differentiation factor 88 (MyD88) (or Toll/IL-1R
(TLR4) activation is capable of upregulating Notch signaling, in- domain containing adaptor inducing IFNβ (TRIF) with the exception of
dependent of microbial interactions [15]. They also found that TLR4 TLR3), which causes downstream signaling to the nuclear factor-κB
and Notch signaling are increased whereas goblet cells and MUC2 ex- (NF-κB) pathway [31–33]. NF-κB is a transcription factor that translo-
pression are decreased in mice and premature infants with NEC [15]. cates to the nucleus and induces an inflammatory response, drawing
Since MUC2 plays an important role in regulating gut barrier integrity, effector cells to the location of the initial pathogen or injury [31]. This
any decrease in MUC2 could potentially increase the risk for NEC. Other conserved response is tightly regulated, yet exacerbated TLR responses
studies have also shown that MUC2 is significantly decreased in infants have been heavily implicated in the pathogenesis of NEC.
with NEC, which can be exacerbated by factors such as immature goblet Within the last few decades, numerous studies have suggested that
cells due to prematurity, increased Notch signaling, or mutations TLR4 activation and signaling is a significant contributor to NEC de-
leading to aberrant goblet cell development [14,16–18]. velopment [34–43]. Since prematurity is a major risk factor in devel-
oping NEC, several studies have investigated differences between the
2.1.3. Paneth cells/antimicrobial peptides preterm and term TLR expression in the intestinal tract. Interestingly,
Paneth cells are significant contributors to the integrity of the in- these studies indicate that there are higher TLR4 expression levels in
testinal epithelium since they produce and secrete peptides that rapidly premature murine and human intestine as compared to full-term con-
kill or inactivate micro-organisms [19–21]. On average, five to 12 Pa- trols [31,44–46]. Investigators have demonstrated that TLR4 activation
neth cells reside near the bottom of the crypt and are renewed every in this immature environment results in increased enterocyte apoptosis,
two to three weeks, allowing them to produce enough antimicrobial reduced enterocyte proliferation and migration, and the eventual
peptides (AMPs) to maintain homeostasis [20,21]. AMPs include a-de- breakdown of the intestinal epithelium [34,36,37,39,40,47]. Once
fensins (human) or cryptidins (mice), lysozyme C, secretory group IIA epithelial integrity is lost, pathogens are able to enter the circulatory
phospholipase A2 (sPLA2), C-type lectins (REG3α in humans, REG3γ in system resulting in systemic inflammation. Moreover, TLR4 activation
mice), and angiogenin 4 (ANG4) in mice [19–21]. In the small intestine, has been shown to reduce endothelial nitric oxide synthase (eNOS) in
Paneth cells secrete these peptides in response to pathogen-associated the intestinal endothelium, causing decreased blood flow and ischemia
molecular patterns (PAMPs), or molecular motifs expressed by classes that may further support the development of NEC [43,48].
of micro-organisms [21]. Studies have also indicated that TLR4 is required for NEC devel-
Although Paneth cells produce AMPs to counteract enteric patho- opment, and is not simply an outcome of the disease. In one such study,
gens, genetic mutations can lead to the disruption of the intestinal TLR4-mutant mice (C3H/HeJ) were protected from developing ex-
epithelium and increase susceptibility of inflammatory bowel diseases perimental NEC, whereas wild-type littermates experienced disruption
(IBD). Crohn's disease (CD) results in ileal inflammation and is asso- of the intestinal epithelium and severe disease [40]. Others utilized
ciated with mutations of NOD2, which is a PRR predominantly present global and intestinal-specific TLR4 gene deletion that demonstrated
in Paneth cells [20,22,23]. NOD2 is a receptor that recognizes muramyl preservation of the intestinal epithelium and downregulation of pro-
dipeptide, which is found in both Gram-positive and Gram-negative inflammatory cytokines [15]. Likewise, it has been shown that, by in-
bacteria. Studies have shown that NOD2-deficient mice have fewer hibiting TLR4 signaling, mice were protected against experimental NEC
cryptidins at both baseline and post-infection as compared to wild-type [34,37,38,46,48]. For example, administration of amniotic fluid [37],
(WT) mice [24]. Additionally, NOD2-deficient mice challenged with breast milk [38], or human milk oligosaccharide 2′-fucosyllactose
Helicobacter hepaticus develop granulomatous lesions in the ileum, (2′FL) [48] attenuated TLR4 signaling, thus significantly decreasing the
which is common in patients with Crohn's disease, and they express pro- severity of experimental NEC in mice. Together, these studies suggest
inflammatory TH1-related genes, such as the cytokine interferon-γ (IFN- an important role for TLR4 in NEC pathogenesis.
γ) and the transcription factor T-bet [20,25]. In human studies, the Other TLRs have also been studied in the context of intestinal injury.
effect of NOD2 loss-of-function mutations has been evaluated to TLR2, which is expressed in enterocytes, immune cells in the lamina
395
propria, and in enteroendocrine cells, induces the production of anti- animals and further studies are needed to confirm their role in NEC.
inflammatory cytokines, such as IL-10, to modulate the immune re-
sponse [49]. However, imbalances in TLR2-mediated NF-κB activation 2.4. Innate lymphoid cells
can lead to pro-inflammatory cytokine production and induce intestinal
injury. Furthermore, TLR2 also supports preservation of tight junctions Within the last decade, a new population of cells in the innate im-
between enterocytes, thus maintaining gut integrity [49]. Studies have mune system, known as innate lymphoid cells (ILCs), have been in-
also demonstrated protection against NEC by modulating the expression creasingly investigated in order to determine their function in im-
of TLR9 [36,39]. By administering CpG-DNA (ligand to TLR9), TLR4- munity. These cells are derived from a common lymphoid progenitor
mediated signaling is modulated and prevents enterocyte apoptosis and and embody the traditional lymphoid cell morphology, yet lack char-
bacterial translocation, thus decreasing experimental NEC severity acteristic receptors and antigen specificity that are typical in adaptive
[39]. Although these studies yield a basic understanding of the inter- lymphocytes [58]. ILCs are largely grouped into three different classes
actions between TLRs and NEC pathogenesis, more studies are needed (subclasses exist among each group; see Ref. [59]), which are defined
to discover additional roles of TLRs during NEC. by cell-surface markers, transcriptional factors, and cytokine expres-
sion; however, each class elicits a distinct immune response [58]. Group
2.2.2. Nucleotide oligomerization domain-like receptors (NLRs) 1 (ILC1s) produces cytokines INF-γ and TNF to protect the intestinal
NLRs are another class of PRRs that recognize PAMPs and damage- epithelium from invading viruses, bacteria, or other intracellular micro-
associated molecular patterns (DAMPs) within the cytoplasm of cells. organisms; Group 2 (ILC2s) produces type-2 cytokines (IL-4, IL-5, IL-9,
NOD1 and NOD2, two members of this class, have been suggested to IL-13) to facilitate the expulsion of parasites from the intestine; and ILC
modulate NEC development via TLR regulation [47,50–52]. NOD1, group 3 (ILC3s) produces lymphotoxin (LT), IL-17A, IL-22, and INF-γ,
expressed by enterocytes, recognizes peptidoglycans found in the cell which help mediate responses to commensal and extracellular microbes
walls of Gram-negative bacteria and activates an immune response [58].
though the Peyer's patches [3,53]. Similarly, NOD2 recognizes in- As investigators have continued to study these cells, research has
tracellular muramyl dipeptide (MDP) that is found in most bacterial cell implicated ILCs in intestinal inflammation [60–63]. For example, some
walls and is expressed by monocytes, dendritic cells, and Paneth cells studies show that patients with CD have markedly increased infiltration
[3,30,53]. Multiple studies have explored the roles played by NOD1 and of ILC1s [60,61]. In a study by Bernink et al., resected bowel from CD
NOD2 in NEC. In a study by Richardson et al., activation of NOD2 by patients contained a significantly higher frequency of ILC1s as com-
MDP decreased TLR4 signaling and decreased NEC severity in mice pared to control bowel, which corresponded with increased INF-γ in
[47]. When MDP was administered, the apoptosis regulatory protein these patients [60]. Murine colitis studies further supported these
SMAC-diablo was downregulated, thus inhibiting TLR4-mediated en- findings and suggested that ILC1 infiltration is a response to CD and
terocyte apoptosis [47]. In another study, investigators found that does not play a causative role [60,61]. ILC3s have also been suggested
NOD2 expression is required for colonization of commensal bacteria to contribute to intestinal inflammation during IBD [62,63]. Buonocore
and suppression of pathogenic bacteria [52]. When H. hepaticus was et al. demonstrated in Rag2−/− mice, which fail to generate B and T-
gavaged to WT and NOD2-deficient mice, NOD2-deficient mice were cells, that ILC3s mount a pro-inflammatory response by producing IL-
unable to clear H. hepaticus infection [52]. In relation to NEC, mutations 17A and IFN-γ in an IL-23-dependent manner [62]. Interestingly, an-
in NOD2 may promote colonization of pathogenic microflora in the other study utilizing the Tbx21−/− Rag2−/− ulcerative colitis (TRUC)
premature intestine and lead to NEC development [52]. Finally, both model of IBD discovered that TNF-α from DCs acted synergistically with
NOD1 and NOD2 activate the NFκB pathway, which, depending on the IL-23 to stimulate ILC3s to produce IL-17A, thus demonstrating a novel
stimulus, could modulate TLR4 expression or act synergistically with interaction between DCs and ILC3s [63]. Taken together, these studies
TLRs [30,47,50]. support the role of ILC1s and ILC3s in the pathogenesis of intestinal
inflammation.
2.3. Neutrophils
3. Adaptive immune system
Neutrophils are major drivers of innate immune responses. They are
the first cells at the site of injury, produce bactericidal compounds, and 3.1. Macrophages
attract and influence other cell types, such as monocytes and dendritic
cells [54]. Their implication in NEC remains to be fully elucidated, yet Macrophages are important effector cells that contribute to the
there are a few studies to note. One study used a Cronobacter sakazakii maintenance of homeostasis as well as initiating an immune response
NEC mouse model to induce NEC in WT mice and mice depleted of during injury. Intestinal macrophages typically reside beneath the
neutrophils, finding an increase in pro-inflammatory cytokines, nitric epithelial layer in the lamina propria, where they are activated upon
oxide, and enterocyte apoptosis in neutrophil-depleted mice (72%) exposure to LPS and IFN-γ and respond by releasing pro-inflammatory
compared to WT mice (40%) [55]. Another study investigated the in- cytokines and nitric oxide [53]. However, macrophage activation is
cidence of neutropenia (neutrophil counts of ≤1000/mL) in small for amplified in the immature intestine, but is downregulated by tumor
gestational age (SGA) human neonates, which suggested that these in- growth factor-β (TGF-β) until the intestine is fully developed [64].
fants have a 4-fold increased risk to develop NEC [56]. These findings Studies related to NEC have observed high intestinal infiltration of
indicate that neutrophils may attenuate the immune response in NEC. macrophages and significantly lower levels of the TGF-β2 isoform,
However, one study demonstrated that overexpression of the bacter- suggesting that an exacerbated macrophage response in the premature
icidal oxidative species produced by neutrophils could contribute to the intestine may play a role in NEC development [64,65].
severity of NEC [57]. In rats with and without neutrophils (depleted In investigating the macrophage response during NEC, studies have
with vinblastine), treatment with platelet activating factor (PAF) and discovered a pathway describing the relationship between TGF-β, mo-
LPS were used to induce NEC [57]. Rats with neutrophils experienced thers against decapentaplegic homolog 7 (Smad7), and NF-κB [66,67].
greater hypotension, reduced intestinal perfusion, and necrotic bowel Along with low levels of TGF-β2 in in-vitro and in-vivo models of NEC,
compared to neutrophil-depleted rats [57]. This suggests that the re- MohanKumar et al. found an increase in expression of Smad7 in mac-
active oxidative metabolites produced by neutrophils may promote rophages and an increased sensitivity to bacterial products, such as LPS
intestinal injury and that the depletion of neutrophils may be protective [64,66,67]. Additionally, Smad7-overexpressing macrophages acti-
in NEC. Nonetheless, the variation in neutrophil function may be con- vated the NF-κB pathway and induced a pro-inflammatory response
founded due to the diverse models used to induce NEC-like disease in [66]. Together, these studies suggest that an imbalance in this
396
macrophage response pathway may influence the inflammatory state nuclear hormone receptor C (RORC) and occludin, which promote early
that occurs in the premature intestine during NEC. pathogen colonization resistence via IL-17 and γδ IELs migration into
the epithelium, respectively [77]. Taken together, these data suggest
3.2. Dendritic cells that γδ IELs are critical for intestinal barrier integrity and that mucosal
γδ IELs may be important in preventing NEC.
Intestinal dendritic cells (DCs) are specialized antigen presenting
cells (APCs) that play a critical role in establishing self-tolerance and 3.5. Regulatory T-cells
maintaining homeostasis in the small intestine [68]. By sampling an-
tigens that pass through the lumen of the gut, DCs can differentiate Regulatory T-cells (Tregs) are CD4+ T-cells that act to suppress
between antigens on pathogenic bacteria from those found on food and excessive immune responses as well as mediate immunological toler-
commensal microbiota. However, investigation of their role in NEC has ance. Interestingly, Tregs that express forkhead box protein 3 (Foxp3)
been fairly limited. In one study, investigators used the opportunistic are present in the developing gut as early as 23 weeks of gestation,
pathogen Cronobacter sakazakii, which may contaminate infant for- suggesting that they may participate in establishing intestinal home-
mula, to induce NEC in newborn mice [69]. When infected with C. sa- ostasis in neonates [79]. Therefore, murine and human studies were
kazakii, a significant number of DCs were recruited to the lamina pro- conducted to explore the immunomodulatory role of Tregs in NEC. In
pria, which resulted in an increased release of cytokines IL-10 and TGF- one such study, Egan et al. induced NEC in WT neonatal mice and
β [69]. With higher levels of cytokines, specifically TGF-β, NEC-related observed significant mucosal barrier damage, increased pro-in-
symptoms such as enterocyte apoptosis and epithelial disruption were flammatory cytokine levels, and a reduction in CD4+ Foxp3+ Tregs
observed [69]. This study concluded that, in a C. sakazakii-induced [34]. They also discovered that the neonatal intestine is rich in signal
model of NEC, DCs contribute to the breakdown of intestinal epithelial transducer and activator of transcription 3 (STAT3), which directs T-
integrity [69]; however, additional studies are necessary to elucidate cell differentiation away from the Treg phenotype [34]. By treating
the role of DCs in premature infants with NEC. neonatal mice with a STAT3 inhibitor, Tregs were restored and NEC
severity decreased [34]. In another study, small intestine was resected
3.3. T-cells from NEC and non-NEC infants and T-cell populations were analyzed
[80]. They found that intestinal resections from NEC patients had a
Historically, neonates were assumed to be deficient in T-cells and significant decrease in Treg proportions compared to other T-cell po-
other adaptive immune cells [70]. However, in 1996, novel studies pulations [80]. Additionally, investigators discovered a pro-in-
investigating the neonatal adaptive immune system demonstrated that flammatory cytokine profile in NEC tissue samples, which correspond
neonatal T-cells were capable of adult-level responses [70–73]. Since to cytokines that have been shown to suppress Tregs [80]. Finally,
then, continued research has shown that neonatal T-cell responses are patients with resolved NEC displayed restoration of the Treg cell po-
increasingly variable and dependent on the timing and conditions pulation, suggesting that Treg depletion is a response to the hostile
surrounding antigen exposure [70]. As such, evidence has emerged microenvironment induced by NEC [80]. Together, these studies sup-
establishing a primary role of T-cells in the development of NEC. For port that Tregs serve a protective role in the fetal and neonatal intes-
the purpose of this review, we focus on T-cell subsets that have been tine.
directly implicated in the pathogenesis of NEC.
3.6. T-helper 17 cells
3.4. Intraepithelial lymphocytes
T-helper cells (TH) are CD4+ T-cells, which, once stimulated by
Intraepithelial lymphocytes (IELs) primarily consist of antigen-ex- APCs, actively participate in cell-mediated immune responses. TH-cells
perienced T-cells that, upon exposure to familiar antigens, release cy- can be further divided into subtypes based on the cytokines secreted
tokines to recruit effector cells to destroy invading pathogens and and the stimuli they respond to; however, we now focus on TH17 cells
growth factors to stimulate epithelial repair [74]. In the gut, IELs are and evaluate their role in NEC. Several studies have demonstrated that
interspersed between epithelial cells of the intestinal epithelium and TH17-derived cytokines can cause intestinal inflammation during ex-
function to elicit and regulate both innate and adaptive responses perimental NEC. For example, Egan et al. had demonstrated a decrease
during infection [74,75]. Thus, mucosal IELs act to preserve the epi- in Tregs in NEC-induced neonatal mice and a significant increase in the
thelial barrier and maintain homeostasis [74,75]. Mucosal IELs are CD4+ TH17 cells, which primarily produce the inflammatory cytokine
highly heterogeneous and are distinguished based on T-cell receptors IL-17A [34]. IL-17A was determined to be a significant contributor to
(TCR) γδ or αβ, as well as on CD8 co-receptor expression [75]. γδ IELs NEC pathogenesis, as demonstrated by the loss of tight junctions be-
are the first T-cell subset present in the intestine during embryogenesis tween intestinal epithelial cells, reduced enterocyte proliferation, and
[76,77]. In the study by Gibbons et al., neonatal γδ IELs were found to an increase in crypt apoptosis after administration to neonatal mice
produce higher levels of cytokines, such as IFN-γ and IL-10, as com- [34]. Additionally, they discovered that the microenvironment of the
pared to neonatal αβ IELs and adult γδ IELs, indicating enhanced ac- premature human intestine expresses high levels of IL-6, IL-22, and
tivity of γδ IELs during early life [76]. Murine studies using the DSS STAT3, all which drive TH17 differentiation [34]. An additional study
model of colitis showed that γδ IELs are important in managing inva- demonstrated that mice subjected to an experimental NEC model ob-
sion of bacteria in the event of mucosal injury [78]. Microarray analysis served severe damage to the mucosal epithelium, primarily due to
exhibited a complex transcriptional response by γδ IELs, including apoptosis of the Lgr5+ intestinal stem cells residing at the bottom of
transcriptional enrichment of cytoprotective properties (i.e. heat shock intestinal crypts [81]. These intestinal stem cells are critical for the
proteins), anti-inflammatory cytokines, and transcription factors that regeneration of other epithelial cell types; therefore, loss of these cells is
promote healing, all of which support epithelial regeneration [78]. detrimental to the integrity of the mucosal barrier [4,81]. Importantly,
Interestingly, this study also demonstrated that commensal bacteria when mice subjected to experimental NEC were treated with all-trans
provide immunomodulatory factors that promote the protective tran- retinoic acid (ATRA), which has been previously shown to induce Treg
scriptional response by γδ IELs during mucosal injury [78]. Further- populations while limiting TH17 differentiation, NEC severity was sig-
more, Weitkamp et al. discovered significantly lower CD8+ γδ IELs in nificantly reduced [34,81]. These studies demonstrate the negative ef-
preterm infants with NEC compared to control infants, suggesting that fects of increased CD4+ TH17 cells on the neonatal small intestine
γδ IELs depletion occurs during the development of NEC [77]. There during NEC, and highlight the use of ATRA as a potential preventative
were also decreases in gene expression for retinoic acid-related orphan strategy or therapeutic for NEC.
397
4. Conclusions function and antigen uptake. Microb Infect 2005;7:997–1004.

[7] Vitale S, Picascia S, Gianfrani C. The cross-talk between enterocytes and in-
traepithelial lymphocytes. Mol Cell Pediatr 2016;3:20.
The complex pathogenesis of NEC has yet to be completely eluci- [8] Hackam DJ, Good M, Sodhi CP. Mechanisms of gut barrier failure in the patho-
dated, and, at least regarding the role of adaptive immunity in NEC, genesis of necrotizing enterocolitis: toll-like receptors throw the switch. Semin
more studies evaluating the differences between NEC in animal models Pediatr Surg 2013;22:76–82.
[9] Mowat AM. Anatomical basis of tolerance and immunity to intestinal antigens. Nat
and human infants are needed. However, recent discoveries have re- Rev Immunol 2003;3:331–41.
sulted in obtaining a more comprehensive understanding of intestinal [10] Deplancke B, Gaskins HR. Microbial modulation of innate defense: goblet cells and
immunology in the premature gut, such as the delicate interplay be- the intestinal mucus layer. Am J Clin Nutr 2001;73:1131S–41S.
[11] Johansson ME, Larsson JM, Hansson GC. The two mucus layers of colon are orga-
tween the intestinal epithelium, commensal bacteria, molecular sig- nized by the MUC2 mucin, whereas the outer layer is a legislator of host-microbial
naling, and innate and adaptive immune cells. Importantly, these interactions. Proc Natl Acad Sci USA 2011;108(Suppl 1):4659–65.
findings have led to testing of novel approaches in NEC prevention such [12] Van der Sluis M, De Koning BA, De Bruijn AC, et al. Muc2-deficient mice sponta-
neously develop colitis, indicating that MUC2 is critical for colonic protection.
as promoting an effective gut barrier, fostering immune surveillance
Gastroenterology 2006;131:117–29.
without injury, and supporting a balanced intestinal microbiome during [13] McDole JR, Wheeler LW, McDonald KG, et al. Goblet cells deliver luminal antigen to
pregnancy and after birth. The latter can be achieved either through CD103+ dendritic cells in the small intestine. Nature 2012;483:345–9.
avoidance of antibiotics, supplementation with milk-derived factors or [14] van Es JH, van Gijn ME, Riccio O, et al. Notch/gamma-secretase inhibition turns
proliferative cells in intestinal crypts and adenomas into goblet cells. Nature
probiotics. Although we have gained valuable insight, continued re- 2005;435:959–63.
search efforts are necessary to define the pathogenesis of NEC and to [15] Sodhi CP, Neal MD, Siggers R, et al. Intestinal epithelial Toll-like receptor 4 reg-
develop more effective strategies to alleviate this disease. ulates goblet cell development and is required for necrotizing enterocolitis in mice.
Gastroenterology 2012;143. 708-18 e1–5.
[16] Hackam DJ, Upperman JS, Grishin A, Ford HR. Disordered enterocyte signaling and
4.1. Practice points intestinal barrier dysfunction in the pathogenesis of necrotizing enterocolitis. Semin
Pediatr Surg 2005;14:49–57.
• NEC is characterized by excessive inflammation likely resulting from [17] Clark JA, Doelle SM, Halpern MD, Saunders TA, Holubec H, Dvorak K, et al.
Intestinal barrier failure during experimental necrotizing enterocolitis: protective
a complex interplay between a host, microbes and nutritionally effect of EGF treatment. Am J Physiol 2006;291:G938–49.
derived antigens. [18] Martin NA, Mount Patrick SK, Estrada TE, et al. Active transport of bile acids de-
• Currently, human milk and a diverse microbiome are the best-stu- creases mucin 2 in neonatal ileum: implications for development of necrotizing
enterocolitis. PLoS One 2011;6:e27191.
died measures to prevent NEC by contributing to a balanced mu- [19] Gallo RL, Hooper LV. Epithelial antimicrobial defence of the skin and intestine. Nat
cosal immune system. Rev Immunol 2012;12:503–16.
[20] Bevins CL, Salzman NH. Paneth cells, antimicrobial peptides and maintenance of
intestinal homeostasis. Nat Rev Microbiol 2011;9:356–68.
4.2. Research directions [21] Elphick DA, Mahida YR. Paneth cells: their role in innate immunity and in-
flammatory disease. Gut 2005;54:1802–9.
• Whereas animal models have been and will remain critical for fur- [22] Hugot JP, Chamaillard M, Zouali H, et al. Association of NOD2 leucine-rich repeat
variants with susceptibility to Crohn's disease. Nature 2001;411:599–603.
ther investigation of the pathogenesis of NEC, the unique feto-ma- [23] Ogura Y, Bonen DK, Inohara N, et al. A frameshift mutation in NOD2 associated
ternal environment with more advanced adaptive immunity at birth, with susceptibility to Crohn's disease. Nature 2001;411:603–6.
make correlation with human cells and tissues necessary. [24] Kobayashi KS, Chamaillard M, Ogura Y, et al. Nod2-dependent regulation of innate
•
and adaptive immunity in the intestinal tract. Science 2005;307:731–4.
Given the important role of the immune system in the pathophy- [25] Biswas A, Liu YJ, Hao L, et al. Induction and rescue of Nod2-dependent Th1-driven
siology of NEC, clinical trials with molecules designed to support granulomatous inflammation of the ileum. Proc Natl Acad Sci USA
intestinal maturation and provide protective immune function or 2010;107:14739–44.
[26] Hartel C, Hartz A, Pagel J, et al. NOD2 Loss-of-function mutations and risks of
inhibit inflammatory processes are needed. necrotizing enterocolitis or focal intestinal perforation in very low-birth-weight
infants. Inflamm Bowel Dis 2016;22:249–56.
Conflicts of interest [27] Zhang C, Sherman MP, Prince LS, et al. Paneth cell ablation in the presence of
Klebsiella pneumoniae induces necrotizing enterocolitis (NEC)-like injury in the
small intestine of immature mice. Dis Models Mechs 2012;5:522–32.
None declared. [28] McElroy SJ, Underwood MA, Sherman MP. Paneth cells and necrotizing en-
terocolitis: a novel hypothesis for disease pathogenesis. Neonatology
2013;103:10–20.
Funding sources
[29] Underwood MA. Paneth cells and necrotizing enterocolitis. Gut Microb
2012;3:562–5.
M.G. is supported by grants K08DK101608 and R03DK111473 from [30] Takeuchi O, Akira S. Pattern recognition receptors and inflammation. Cell
the National Institutes of Health, March of Dimes Foundation Grant No. 2010;140:805–20.
[31] Lu P, Sodhi CP, Hackam DJ. Toll-like receptor regulation of intestinal development
5-FY17-79, and the Children's Discovery Institute of Washington and inflammation in the pathogenesis of necrotizing enterocolitis. Pathophysiology
University and St. Louis Children's Hospital. 2014;21:81–93.
[32] Takeda K, Akira S. Roles of Toll-like receptors in innate immune responses. Gene
Cell 2001;6:733–42.
Acknowledgment [33] Barton GM, Kagan JC. A cell biological view of Toll-like receptor function: reg-
ulation through compartmentalization. Nat Rev Immunol 2009;9:535–42.
We apologize to the authors whose work we were not able to in- [34] Egan CE, Sodhi CP, Good M, et al. Toll-like receptor 4-mediated lymphocyte influx
induces neonatal necrotizing enterocolitis. J Clin Invest 2016;126:495–508.
clude due to space limitations. [35] Neal MD, Leaphart C, Levy R, et al. Enterocyte TLR4 mediates phagocytosis and
translocation of bacteria across the intestinal barrier. J Immunol 2006;176:3070–9.
References [36] Afrazi A, Branca MF, Sodhi CP, et al. Toll-like receptor 4-mediated endoplasmic
reticulum stress in intestinal crypts induces necrotizing enterocolitis. J Biol Chem
2014;289:9584–99.
[1] Papillon S, Castle SL, Gayer CP, Ford HR. Necrotizing enterocolitis: contemporary [37] Good M, Siggers RH, Sodhi CP, et al. Amniotic fluid inhibits Toll-like receptor 4
management and outcomes. Adv Pediatr 2013;60:263–79. signaling in the fetal and neonatal intestinal epithelium. Proc Natl Acad Sci USA
[2] Nino DF, Sodhi CP, Hackam DJ. Necrotizing enterocolitis: new insights into pa- 2012;109:11330–5.
thogenesis and mechanisms. Nat Rev Gastroenterol Hepatol 2016;13:590–600. [38] Good M, Sodhi CP, Egan CE, et al. Breast milk protects against the development of
[3] Santaolalla R, Abreu MT. Innate immunity in the small intestine. Curr Opin necrotizing enterocolitis through inhibition of Toll-like receptor 4 in the intestinal
Gastroenterol 2012;28:124–9. epithelium via activation of the epidermal growth factor receptor. Mucosal
[4] van der Flier LG, Clevers H. Stem cells, self-renewal, and differentiation in the in- Immunol 2015;8:1166–79.
testinal epithelium. Annu Rev Physiol 2009;71:241–60. [39] Gribar SC, Sodhi CP, Richardson WM, et al. Reciprocal expression and signaling of
[5] Miron N, Cristea V. Enterocytes: active cells in tolerance to food and microbial TLR4 and TLR9 in the pathogenesis and treatment of necrotizing enterocolitis.
antigens in the gut. Clin Exp Immunol 2012;167:405–12. J Immunol 2009;182:636–46.
[6] Snoeck V, Goddeeris B, Cox E. The role of enterocytes in the intestinal barrier [40] Leaphart CL, Cavallo J, Gribar SC, et al. A critical role for TLR4 in the pathogenesis
398
of necrotizing enterocolitis by modulating intestinal injury and repair. J Immunol unique subset of IL-12- and IL-15-responsive IFN-gamma-producing cells. Immunity
2007;179:4808–20. 2013;38:769–81.
[41] Neal MD, Sodhi CP, Jia H, et al. Toll-like receptor 4 is expressed on intestinal stem [62] Buonocore S, Ahern PP, Uhlig HH, et al. Innate lymphoid cells drive interleukin-23-
cells and regulates their proliferation and apoptosis via the p53 up-regulated dependent innate intestinal pathology. Nature 2010;464:1371–5.
modulator of apoptosis. J Biol Chem 2012;287:37296–308. [63] Powell N, Walker AW, Stolarczyk E, et al. The transcription factor T-bet regulates
[42] Sodhi CP, Shi XH, Richardson WM, et al. Toll-like receptor-4 inhibits enterocyte intestinal inflammation mediated by interleukin-7 receptor+ innate lymphoid cells.
proliferation via impaired beta-catenin signaling in necrotizing enterocolitis. Immunity 2012;37:674–84.
Gastroenterology 2010;138:185–96. [64] Maheshwari A, Kelly DR, Nicola T, et al. TGF-beta2 suppresses macrophage cyto-
[43] Yazji I, Sodhi CP, Lee EK, et al. Endothelial TLR4 activation impairs intestinal mi- kine production and mucosal inflammatory responses in the developing intestine.
crocirculatory perfusion in necrotizing enterocolitis via eNOS-NO-nitrite signaling. Gastroenterology 2011;140:242–53.
Proc Natl Acad Sci USA 2013;110:9451–6. [65] MohanKumar K, Kaza N, Jagadeeswaran R, et al. Gut mucosal injury in neonates is
[44] Nanthakumar N, Meng D, Goldstein AM, et al. The mechanism of excessive in- marked by macrophage infiltration in contrast to pleomorphic infiltrates in adult:
testinal inflammation in necrotizing enterocolitis: an immature innate immune re- evidence from an animal model. Am J Physiol 2012;303:G93–102.
sponse. PLoS One 2011;6:e17776. [66] MohanKumar K, Namachivayam K, Chapalamadugu KC, et al. Smad7 interrupts
[45] Fusunyan RD, Nanthakumar NN, Baldeon ME, Walker WA. Evidence for an innate TGF-beta signaling in intestinal macrophages and promotes inflammatory activa-
immune response in the immature human intestine: toll-like receptors on fetal tion of these cells during necrotizing enterocolitis. Pediatr Res 2016;79:951–61.
enterocytes. Pediatr Res 2001;49:589–93. [67] Namachivayam K, Blanco CL, MohanKumar K, et al. Smad7 inhibits autocrine ex-
[46] Neal MD, Sodhi CP, Dyer M, et al. A critical role for TLR4 induction of autophagy in pression of TGF-beta2 in intestinal epithelial cells in baboon necrotizing en-
the regulation of enterocyte migration and the pathogenesis of necrotizing enterocolitis. Am J Physiol 2013;304:G167–80.
terocolitis. J Immunol 2013;190:3541–51. [68] Steimle A, Frick JS. Molecular mechanisms of induction of tolerant and tolerogenic
[47] Richardson WM, Sodhi CP, Russo A, et al. Nucleotide-binding oligomerization do- intestinal dendritic cells in mice. J Immunol Res 2016;2016:1958650.
main-2 inhibits toll-like receptor-4 signaling in the intestinal epithelium. [69] Emami CN, Mittal R, Wang L, Ford HR, Prasadarao NV. Recruitment of dendritic
Gastroenterology 2010;139:904–17. 17 e1–6. cells is responsible for intestinal epithelial damage in the pathogenesis of necro-
[48] Good M, Sodhi CP, Yamaguchi Y, et al. The human milk oligosaccharide 2′-fuco- tizing enterocolitis by Cronobacter sakazakii. J Immunol 2011;186:7067–79.
syllactose attenuates the severity of experimental necrotising enterocolitis by en- [70] Adkins B, Leclerc C, Marshall-Clarke S. Neonatal adaptive immunity comes of age.
hancing mesenteric perfusion in the neonatal intestine. Br J Nutr Nat Rev Immunol 2004;4:553–64.
2016;116:1175–87. [71] Sarzotti M, Robbins DS, Hoffman PM. Induction of protective CTL responses in
[49] Cario E. Barrier-protective function of intestinal epithelial Toll-like receptor 2. newborn mice by a murine retrovirus. Science 1996;271:1726–8.
Mucosal Immunol 2008;1(Suppl 1):S62–6. [72] Ridge JP, Fuchs EJ, Matzinger P. Neonatal tolerance revisited: turning on newborn
[50] Strober W, Murray PJ, Kitani A, Watanabe T. Signalling pathways and molecular T cells with dendritic cells. Science 1996;271:1723–6.
interactions of NOD1 and NOD2. Nat Rev Immunol 2006;6:9–20. [73] Forsthuber T, Yip HC, Lehmann PV. Induction of TH1 and TH2 immunity in neo-
[51] Le Mandat Schultz A, Bonnard A, Barreau F, et al. Expression of TLR-2, TLR-4, natal mice. Science 1996;271:1728–30.
NOD2 and pNF-kappaB in a neonatal rat model of necrotizing enterocolitis. PLoS [74] Cheroutre H, Lambolez F, Mucida D. The light and dark sides of intestinal in-
One 2007;2:e1102. traepithelial lymphocytes. Nat Rev Immunol 2011;11:445–56.
[52] Petnicki-Ocwieja T, Hrncir T, Liu YJ, et al. Nod2 is required for the regulation of [75] Sheridan BS, Lefrancois L. Intraepithelial lymphocytes: to serve and protect. Curr
commensal microbiota in the intestine. Proc Natl Acad Sci USA 2009;106:15813–8. Gastroenterol Rep 2010;12:513–21.
[53] McElroy SJ, Weitkamp JH. Innate immunity in the small intestine of the preterm [76] Gibbons DL, Haque SF, Silberzahn T, et al. Neonates harbour highly active gam-
infant. NeoReviews 2011;12:e517–26. madelta T cells with selective impairments in preterm infants. Eur J Immunol
[54] Nathan C. Neutrophils and immunity: challenges and opportunities. Nat Rev 2009;39:1794–806.
Immunol 2006;6:173–82. [77] Weitkamp JH, Rosen MJ, Zhao Z, et al. Small intestinal intraepithelial
[55] Emami CN, Mittal R, Wang L, Ford HR, Prasadarao NV. Role of neutrophils and TCRgammadelta+ T lymphocytes are present in the premature intestine but se-
macrophages in the pathogenesis of necrotizing enterocolitis caused by Cronobacter lectively reduced in surgical necrotizing enterocolitis. PLoS One 2014;9:e99042.
sakazakii. J Surg Res 2012;172:18–28. [78] Ismail AS, Behrendt CL, Hooper LV. Reciprocal interactions between commensal
[56] Christensen RD, Yoder BA, Baer VL, Snow GL, Butler A. Early-onset neutropenia in bacteria and gamma delta intraepithelial lymphocytes during mucosal injury.
small-for-gestational-age infants. Pediatrics 2015;136:e1259–67. J Immunol 2009;182:3047–54.
[57] Musemeche C, Caplan M, Hsueh W, Sun X, Kelly A. Experimental necrotizing en- [79] Weitkamp JH, Rudzinski E, Koyama T, et al. Ontogeny of FOXP3(+) regulatory T
terocolitis: the role of polymorphonuclear neutrophils. J Pediatr Surg cells in the postnatal human small intestinal and large intestinal lamina propria.
1991;26:1047–9. discussion 9–50. Pediatr Dev Pathol 2009;12:443–9.
[58] Artis D, Spits H. The biology of innate lymphoid cells. Nature 2015;517:293–301. [80] Weitkamp JH, Koyama T, Rock MT, et al. Necrotising enterocolitis is characterised
[59] Gury-BenAri M, Thaiss CA, Serafini N, et al. The spectrum and regulatory landscape by disrupted immune regulation and diminished mucosal regulatory (FOXP3)/ef-
of intestinal innate lymphoid cells are shaped by the microbiome. Cell fector (CD4, CD8) T cell ratios. Gut 2013;62:73–82.
2016;166:1231–1246 e13. [81] Nino DF, Sodhi CP, Egan CE, et al. Retinoic acid improves incidence and severity of
[60] Bernink JH, Peters CP, Munneke M, et al. Human type 1 innate lymphoid cells necrotizing enterocolitis by lymphocyte balance restitution and repopulation of
accumulate in inflamed mucosal tissues. Nat Immunol 2013;14:221–9. LGR5+ intestinal stem cells. Shock 2017;47:22–32.
[61] Fuchs A, Vermi W, Lee JS, et al. Intraepithelial type 1 innate lymphoid cells are a
399

Necrotizing enterocolitis: The intestinal microbiome, metabolome and T

inflammatory mediators
Josef Neua,∗, Mohan Pammib
a
Section of Neonatology, Department of Pediatrics, University of Florida, Gainesville, FL, USA
b
Section of Neonatology, Department of Pediatrics, Baylor College of Medicine and Texas Children's Hospital, Houston, TX, USA
Keywords: Necrotizing enterocolitis (NEC) is a disease of preterm infants and associated with significant mortality and
Microbiome morbidity. Although the pathogenesis of NEC is not clear, microbial dysbiosis, with a bloom of the phylum
Intestinal Proteobacteria, has been reported. Antibiotics and the use of H2 blockers, which affect the gut microbiome, are
Preterm associated with increased incidence of NEC. In association with dysbiosis, inflammatory processes are upregu-
Necrotizing enterocolitis
lated with increased Toll-like receptor signaling, leading to translocation of nuclear factor kappa-β, a tran-
scription factor that induces transcription of various pro-inflammatory cytokines and chemokines. Microbial
metabolites, short chain fatty acids including acetate and butyrate, may modulate immunity, inflammation,
intestinal integrity and regulate transcription by epigenetic mechanisms. Evaluation of microbiome and meta-
bolome may provide biomarkers for early diagnosis of NEC and microbial therapeutic approaches to correct
microbial dysbiosis.
1. Necrotizing enterocolitis outnumber human cells and genes. Host–microbe interactions on in-
testinal and respiratory mucosal surfaces play a major role in the pre-
Necrotizing enterocolitis (NEC) is a major cause of mortality in vention of infections and development of immunity. Intestinal micro-
preterm infants who survive the first few days after birth [1,2]. NEC is biota, the best studied, has a bacterial load in the magnitude of 1014
seen in 7% of very low birth weight infants (birth weight < 1500 g) and bacteria/mm3, which makes it the most densely colonized surface of the
up to 5% of admissions to the neonatal intensive care unit (NICU) [2,3]. human body. Bacteroidetes represent the most abundant phylum, fol-
NEC is associated with a mortality of 15–30% and a significant long- lowed by Firmicutes [11,12]. In healthy adults, Bacteroides, Faecali-
term neurodevelopmental morbidity but we have made little progress bacterium, and Bifidobacterium are the most prevalent genera [12].
for 60 years [3–5]. The pathogenesis of NEC is not clear but microbial The lower respiratory tract is one of the least-populated surfaces of the
dysbiosis, formula feeding and excessive inflammation have been im- human body with an estimated number of 10–100 bacteria per 1000
plicated [3]. Furthermore, a clear definition of this disease remains human cells [13]. The “core microbiota” of healthy individuals consists
elusive and it likely represents several different diseases with the final mainly of genera Pseudomonas, Streptococcus, Prevotella, Fuso-
outcome of intestinal injury or necrosis [6]. bacteria, Veillonella, Haemophilus, Neisseria, and Porphyromonas
Postmenstrual age is correlated with the development of NEC, the [14–16].
peak age being 29–31 weeks (peak 31 weeks) (based on last menstrual
period or conception) [7–9]. Fig. 1 summarizes several data sets on the 2.1. Normal development of a preterm intestinal microbiome
post-conceptual timing of NEC (Fig. 1) [10]. Prematurity is the major
risk factor, but, since not all preterm infants develop this disease, sev- Before birth, the fetus is bathed by the amniotic fluid that may not
eral other environmental factors including intestinal dysbiosis may be be sterile [17]. Rather, the fetal–maternal unit is constantly exposed to
involved. microbes and microbial metabolites that may originate from the vagina,
and mother's gastrointestinal tract including the mouth [18]. Recent
2. Microbiome and human health studies suggest that amniotic fluid contains microbes that can most
readily be detected with non-culture based techniques and that the
The microbial communities on the surface of the human body placenta harbors microbial DNA [17,18]. The fetus ingests large
∗
Corresponding author. Section of Neonatology, Department of Pediatrics, University of Florida, 1600 SW Archer Road, Gainesville, FL, 32607, USA.
E-mail address: neuj@peds.ufl.edu (J. Neu).
J. Neu, M. Pammi Seminars in Fetal and Neonatal Medicine 23 (2018) 400–405
bacterial species, less diversity and increased proportion of potential

pathogens [31,32]. The microbial dysbiosis theory of NEC is supported
by the fact that NEC cannot be produced in germ-free animals [26,33].
It is also suggested by an association between early antibiotic use and
NEC [34,35].
2.3. The intestinal microbial milieu prior to the diagnosis of NEC
Early studies utilizing 16S rDNA for microbiome evaluation sug-

gested that relative increases in the phylum Proteobacteria when
compared to other major microbial phyla such as Firmicutes,
Bacteroides and Negativicutes are associated with the development of
NEC [36,37]. This intriguing relationship suggests that these may be-
Fig. 1. Post-conceptual timing of necrotizing enterocolitis (NEC) diagnosis.
CGA, corrected gestational age.
come targets of microbial intervention against NEC, but true causality
[y-Axis label: ‘No. of NEC cases’]. has not yet been demonstrated and will be a requisite in the determi-
nation of whether a certain individual microbe or microbial pattern is
responsible for the disease. The microbial version of Koch's postulates
quantities, up to 150 mL/kg/d, of amniotic fluid and the fetal gastro-
needs to be fulfilled for causality to be established.
intestinal tract is exposed to large number of microbes and microbial
There are issues with 16S rRNA gene sequencing in the evaluation of
components during gestation. Several studies demonstrate that the
the human microbiome. Taxonomic resolution is limited, and strain-
meconium contains microbial DNA as well as live microbes [19,20].
level resolution and tracking are not possible. The metabolic functions
Studies suggest that the composition of these microbes in amniotic
of the microbiome cannot be assessed. In contrast, metagenomics or
fluid, meconium and placenta differ depending on gestational age of
whole genome shot-gun sequencing can provide strain-level resolution.
delivery. Meconium microbes most likely reflect the in-utero rather
In addition to understanding the microbial community structure, the
than the extrauterine environment. The microbial milieu in the meco-
metabolic potential of the microbial communities can also be assessed
nium of babies may determine the development of sepsis and NEC [19].
[38,39]. 16S rRNA gene evaluation is based on existing information of
After birth, the preterm intestinal microbiome varies with post-
conserved sequences, and unusual sequences are missed. The metage-
menstrual age and undergoes a patterned progression and increase in
nomic approach enables an unbiased evaluation of the microbial
anaerobes (around 36 weeks corrected gestational age (CGA)), in-
communities even those with unusual RNA sequences or plasmids.
dicating maturation of the microbiome [21]. An enumeration of mi-
Meta-genomic approach may also be useful in the diagnosis of a disease
crobes of stools from preterm infants using 16SrRNA sequencing, has
due to an unusual or unsuspected organism [40]. Metagenomic ap-
determined that the microbiota progresses through a succession of the
proaches are labor intensive and much more expensive than the 16S
bacterial classes from Bacilli, to Gammaproteobacteria, to Clostridia
rRNA gene sequencing methods. Metagenomic studies on NEC are
[21]. By the time the infants approach 33–36 weeks post conceptual
scarce although some small-sample studies have been published
age, anaerobes constitute a large component of the gut colonization.
[41,42]. Raveh-Sidka et al. studied the stool microbiome in preterm
infants (five with NEC and five controls) by the meta-genomic approach
2.2. Microbial dysbiosis and NEC and reported that although many microbial species were found in many
infants, the strains colonizing each infant were distinct, indicating a
The concept of NEC associated with inappropriate colonization of paucity of shared gut colonization [41]. In another meta-genomic study
the premature intestine was introduced by Claude and Walker [22]. of the early intestinal microbiome of 144 preterm and 22 term infants,
This was because “epidemics” of NEC had occurred, but no pathogen NEC and mortality were associated with uropathogenic E. coli coloni-
had been associated with NEC. The authors posited that, rather than a zation [43].
direct infection, NEC may be the result of a secondary inflammation in La Rosa et al. observed an orchestrated patterned progression of gut
response to microbial colonization or infection. Histologic sections of microbiota towards an abundance of Clostridia with increasing gesta-
intestine from infants with NEC often showed evidence of necrosis and tional age [21]. Antibiotics, feeding or mode of delivery caused abrupt
inflammation in addition to bacterial overgrowth. They suggested a shifts in microbiota but did not change this predestined progression.
cycle of microbe-induced neutrophil activation leading to inflammatory Other studies comparing gut microbiota in infants with NEC and con-
mediator release, vasoconstriction, and disruption of the intestinal trols have discussed a Proteobacteria bloom positively associated [44]
barrier that was initially caused by interaction of a “dysbiotic” in- and anaerobes (especially Negativicutes) negatively associated with
testinal microbiota and an immature intestine. They also suggested that NEC [37]. In our meta-analyses of microbiome studies in preterm in-
various factors such as breastfeeding could reduce colonization by pa- fants – NEC versus controls who did not develop NEC – we found
thogenic organisms but induce colonization by commensal organisms, gradual shifts in the relative abundances of multiple phyla at about 27
which in turn modulate inflammatory reactions and decreased in- weeks CGA, where decreased abundances of Firmicutes and Bacter-
testinal injury. Other studies have implicated microbial dysbiosis in the oidetes and increased abundances of Proteobacteria preceded the di-
gastrointestinal tract and an exaggerated inflammatory response in the agnosis of NEC [10] (Fig. 2).
development of NEC [23–25]. Immune dysregulation in association
with microbial dysbiosis has also been implicated in the pathogenesis of 2.4. Factors that alter the microbiome and may predispose to NEC
NEC. Excessive Toll-like receptor-4 (TLR4) signaling in response to li-
popolysaccharide (LPS) [26,27] and an exaggerated inflammatory re- Several studies evaluating antibiotic use in preterm infants suggest
sponse [24,25] in preterm infants have been reported. an increased odds ratio in the development of NEC, which is directly
Prior to the last decade, several studies have searched for a specific related to the days of antibiotic usage [35,45,46]. Studies by La Rosa
microbial origin of NEC using culture-based techniques [28–30]. Re- et al. and Murgas et al. suggest a relationship between antibiotic use in
cently, culture-independent techniques have been employed to evaluate preterm infants, the microbial composition of the gastrointestinal tract,
the microbial environment. Studies have reported dysbiotic microbial and the subsequent development of NEC [21,47]. Additional studies
patterns rather than a single pathogen causing NEC [10]. Compared to suggest that the acid–base environment of the gastrointestinal tract may
term infants, the intestinal microbiota in preterm infants has fewer also play a significant role in the development of NEC [48]. This may
401
Fig. 2. Microbial profiles in necrotizing enterocolitis (NEC).
402
also relate to the types of microbe present in the gastrointestinal tract mice and TLR4 knockout mice are protected from NEC [61–64]. It is
because they respond differently to the acid–base environment in terms possible that inhibition of the TLR4 pathway may be a novel strategy in
of their growth. H2 blockers favor the proliferation of the Proteo- the treatment or prevention of NEC.
bacteria over Firmicutes in the gastrointestinal tract [49]. This re-
lationship is highly intriguing since the Proteobacteria to Firmicute 4. Metabolome in necrotizing enterocolitis
ratios also are altered in those infants who subsequently develop NEC
[10]. Animal studies in rabbits who were fed acidified milk had de- Microbiota-derived metabolites, short chain fatty acids (SCFAs),
creased gut colonization and translocation, emphasizing the protective butyrate, propionate and acetate induce IL-18 production from the in-
effects of gastric acidity on microbial colonization of the gut [50,51]. testinal epithelial cells (IECs) through activation of NOD-like family,
Human studies in which a formula was acidified resulted in a lower receptors (NLRs) [65]. Acetate produced by Bifidobacteria promotes
incidence of NEC [52]. Retrospective database reviews and secondary epithelial cell barrier function by inducing an anti-apoptotic response in
analysis of large multi-center studies show that the use of antacids, both the IECs. Microbiota-derived sphingolipids presented on CD1d by
H2 blockers and proton pump inhibitors increase the risk for the de- dendritic cells inhibit colonic invariant natural killer T-cell develop-
velopment of NEC [53,54]. Bovine lactoferrin supplementation miti- ment. Thus microbiota and their metabolites mediate immune response
gates this effect [53]. via the IECs and immune cells [66,67]. Germ-free mice have lower le-
The well-known phenomena that babies receiving their own mo- vels of SCFAs compared to conventionally housed animals [68,69].
ther's milk have a decreased incidence of NEC is likely to relate to many Butyrate also is a major fuel for colonocytes and is a major stimulus for
factors in the milk, including its macronutrient composition, the com- tight junction formation and integrity [70]. SCFAs have also been
position of polyunsaturated fatty acids, lactoferrin, various immune shown to stimulate histone acetylation of FoxP3 (forehead box P3)
cells, immunoglobulins, and microbes. The fact that human milk con- locus on naïve CD4+ T-cells, increase FoxP3 expression and promote
tains microbes that appear to be personalized for each mother's own differentiation of Tregs (anti-inflammatory effects) [65,68,71]. Faecali-
infant [55] is a fascinating phenomenon that may also relate to the bacterium prausnitzii and Eubacterium rectale/Roseburia species (Firmi-
decreased risk of NEC in babies fed their own mothers' milk. The fact cutes) are major contributors of butyrate which regulates gene ex-
that donor milk is usually pasteurized and thus devoid of cellular ma- pression by histone modifications [72]. LPS – an inflammatory marker
terial as well as live microbes could be highly germane in this regard. for cardiovascular disease – may also have a role in epigenetic regula-
Studies utilizing only donor milk in comparison to commercial formula tion of intestinal and immune cells [73].
have not overwhelmingly favored the donor human milk and the pre-
vention of NEC. Some of the studies included in meta-analyses of donor 4.1. Clinical studies of the metabolome in neonates
milk versus formula are nearly 30 years old and it is difficult to discern
the criteria utilized in these studies to diagnose NEC in these older ‘Metabolomics’ was coined in 1998 by Oliver et al. [74] and is the
studies. science of detecting small molecules, the result of metabolic pathways
from biological specimens such as plasma, serum, urine and tissues, and
3. Microbiome immunity and inflammation is the latest of the ‘omics’ technologies. Metabolomics detects the pro-
ducts of the metabolic pathways in an organism which may be useful in
3.1. Microbial–host interactions and innate immunity, inflammation diagnosis, prediction, prognosis or assigning disease status (biomarker
detection). Metabolomics allows identification of distinct patterns of
The relationship between microbes and the innate immune system small molecules generated during both host and microbial cellular
in the gastrointestinal tract has been reviewed [56,57]. Intestinal cells metabolism and may be useful in searching for biomarkers of micro-
including the intestinal epithelium harbor receptors to microbial com- biome patterns and dysbiosis [75–79]. Metabolite patterns are dy-
ponents, TLRs, which play a major role in innate immunity [26,58]. namic, changing with gestational age, time or disease process and at
Activation of these Toll-like receptors results in signaling cascades that any time give us a snapshot of the metabolic milieu of the organism.
induce nuclear translocation of nuclear factor kappa-β (NFKβ), a tran- The complexity and the numerous metabolites to be measured need
scription factor that induced transcription of various pro- and anti-in- sophisticated analytical techniques. Nuclear magnetic resonance (NMR)
flammatory cytokines and chemokines [25]. For example, activation of spectroscopy and mass spectrometry (MS) are the techniques most
the chemokine interleukin (IL)-8 attracts neutrophils to trigger areas of widely employed.
the intestine, where they undergo phagocytosis, and other in- Very few studies have performed on metabolomic patterns in NEC
flammatory responses which subsequently cause microvascular con- patients, compared to controls or healthy preterm infants [80–84].
striction, ischemia and injury to the intestine [59]. Certain microbes Although there is no unifying metabolomic signature in NEC, these
may also affect dendritic cells, which are of the monocyte/macrophage studies have reported interesting results. A multi-center evaluation of
lineage. These cells can traverse the intestinal epithelium, grasp anti- the gut microbiome and metabolome named ‘Mechanisms affecting the
gens from the lumen, carry them to the sub-epithelium and act as an- gut of preterm infants in enteral feeding trials (MAGPIE)’ study, funded
tigen presenting cells to undifferentiated lymphocytes in the intestinal by the UK NIHR Efficacy and Mechanistic Evaluation program, is in
lamina propria. Depending on the surface receptors that are stimulated progress [85].
by the distinct microbial patterns to which these cells are exposed, in- Morrow et al. evaluated urinary metabolome patterns in association
teraction between dendritic cells may result in the differentiation of T- with gut dysbiosis preceding NEC and found that three urinary meta-
cells into effector versus tolerogenic cell types. The precise mechanisms bolites differed in those with NEC compared to those without the dis-
of these interactions as they relate to NEC are currently poorly under- ease [81]. Alanine in the urine was directly correlated with the relative
stood. abundance of Firmicutes (P = 0.009), and was inversely correlated
In our meta-analyses of microbiome studies in NEC, Proteobacteria with the relative abundance of both Proteobacteria (P = 0.027) and
abundances increase in babies who develop NEC compared to controls Propionibacterium (P = 0.015) [81]. Histidine was not independently
[10]. LPS produced by Proteobacteria is recognized by TLR4 in the associated with microbial community characteristics, but ratio of ala-
intestine, which triggers intestinal inflammation and enterocyte injury nine to histidine was positively associated with overall NEC (Krus-
that may lead to the development of NEC in preterm infants [60]. TLR4 kal–Wallis, P = 0.001) and inversely associated with the relative
expression is higher in the intestinal tract of preterm neonates when abundance of Propionibacterium (Spearman's rho = −0.57,
compared to terms, and thus may regulate the balance between in- P = 0.002).
testinal repair and injury [26]. It is interesting to note that TLR4 mutant Wilcock et al. described metabolomic differences in serum between
403
preterm infants who developed NEC (five preterm Furthermore, US Food and Drug Administration-based drug standards
infants < 28 weeks GA) and those who did not (seven should be utilized if these agents are to be used for prevention of NEC.
preterm < 29 weeks GA and eight term infants) [82]. Serum samples in Whether fecal microbial transplant-like approaches such as those being
the first week of life in these preterm infants did not predict who would used in the treatment of Clostridium difficile infections may be of benefit
later develop NEC. However, when the infants were on complete in the prevention of NEC could be a basis of future investigation.
feeding, analyses revealed differences in 12 metabolites related to Transfaunation of donor milk using small amounts of the baby's own
carbohydrate, lipid metabolism and intracellular signaling, between mother's milk may also be a means to “personalize” the donor milk to
NEC and control infants. Based on the metabolites identified, an in- provide a means to protect the infant's intestinal tract from the mi-
hibition of the pro-insulin and upregulation of IL-1β were correlated crobial environment that predisposes to NEC [86].
with the development of NEC.
Stewart et al. studied a large cohort of preterm infants (641 samples Conflicts of interest
from 35 infants from a cohort of > 300 infants) and compared preterm
infants with NEC (n = 7) to those without (n = 28) by 16S rRNA gene None declared.
profiling and metabolomic analysis of stools [84]. A subset of 75
samples (n = 16 patients) were also subjected to metabolomic analysis Funding sources
by ultra-performance liquid chromatography mass spectrometry. At the
time of NEC diagnosis, five metabolites were significantly different in None.
preterm infants with NEC, two metabolites from C21-steroid hormone
biosynthesis and linoleate metabolism pathways, two metabolites from References
the linoleate metabolism pathway, and one metabolite from the leu-
kotriene metabolism and prostaglandin formation from arachidonate [1] Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011;364:255–64.
pathway. Stewart et al. also reported a proteomic and metabolomic [2] Fanaroff AA, Stoll BJ, Wright LL, et al. Trends in neonatal morbidity and mortality
for very low birthweight infants. Am J Obstet Gynecol 2007;196:147. e1–8.
analysis of the serum in preterm infants who developed NEC with [3] Lin PW, Stoll BJ. Necrotising enterocolitis. Lancet 2006;368:1271–83.
controls but did not identify any protein or metabolite associated with [4] Stoll BJ, Hansen NI, Adams-Chapman I, et al. Neurodevelopmental and growth
the development of NEC [80]. impairment among extremely low-birth-weight infants with neonatal infection.
J Am Med Assoc 2004;292:2357–65.
De Magistris et al. compared six infants who developed NEC with 12 [5] Bell MJ, Ternberg JL, Feigin RD, et al. Neonatal necrotizing enterocolitis.
controls by NMR-based urinary metabolic profiles. The investigators Therapeutic decisions based upon clinical staging. Ann Surg 1978;187:1–7.
found increased urinary gluconic acid in NEC patients. Gluconic acid [6] Gordon PV, Swanson JR, MacQueen BC, Christensen RD. A critical question for NEC
researchers: can we create a consensus definition of NEC that facilitates research
can be a byproduct of calcium gluconate administration or a byproduct progress? Semin Perinatol 2017;41:7–14.
of bacterial metabolism by alternate metabolic pathways [7] Yee WH, Soraisham AS, Shah VS, Aziz K, Yoon W, Lee SK. Incidence and timing of
(Entner–Doudoroff pathway). It is possible that production of gluconic presentation of necrotizing enterocolitis in preterm infants. Pediatrics
2012;129:e298–304.
acid can favor bacteria that utilize gluconic acid but are more virulent
[8] Gonzalez-Rivera R, Culverhouse RC, Hamvas A, Tarr PI, Warner BB. The age of
[83]. necrotizing enterocolitis onset: an application of Sartwell's incubation period model.
There is no unifying hypothesis related to altered metabolism in J Perinatol 2011;31:519–23.
NEC from published studies, and the use of metabolomic techniques in [9] Gordon PV, Clark R, Swanson JR, Spitzer A. Can a national dataset generate a no-
mogram for necrotizing enterocolitis onset? J Perinatol 2014;34:732–5.
understanding NEC pathogenesis is at its infancy. Integration of all [10] Pammi M, Cope J, Tarr PI, et al. Intestinal dysbiosis in preterm infants preceding
omics platforms is needed to understand the pathophysiology in NEC. necrotizing enterocolitis: a systematic review and meta-analysis. Microbiome
2017;5:31.
[11] Eckburg PB, Bik EM, Bernstein CN, et al. Diversity of the human intestinal microbial
5. Future directions flora. Science 2005;308:1635–8.
[12] Robles Alonso V, Guarner F. Linking the gut microbiota to human health. Br J Nutr
For the prediction and diagnosis of NEC, diagnostic biomarkers in- 2013;109(Suppl 2):S21–6.
[13] Sze MA, Dimitriu PA, Hayashi S, et al. The lung tissue microbiome in chronic ob-
clude those that may not be specific for NEC such as the C-reactive structive pulmonary disease. Am J Respir Crit Care Med 2012;185:1073–80.
protein, white blood cell count, and platelet count. However, more [14] Erb-Downward JR, Thompson DL, Han MK, et al. Analysis of the lung microbiome
specific biomarkers are being developed which include tight junction in the “healthy” smoker and in COPD. PLoS One 2011;6:e16384.
[15] Charlson ES, Bittinger K, Haas AR, et al. Topographical continuity of bacterial
proteins such as claudin-3, intestinal epithelial cell proteins such as
populations in the healthy human respiratory tract. Am J Respir Crit Care Med
intestinal fatty acid binding protein (IFAB), both of which can be 2011;184:957–63.
analyzed non-invasively in the urine. More predictive biomarkers using [16] Charlson ES, Diamond JM, Bittinger K, et al. Lung-enriched organisms and aberrant
bacterial and fungal respiratory microbiota after lung transplant. Am J Respir Crit
the proteomic, metabolomic and genomic platforms are being eval-
Care Med 2012;186:536–45.
uated. A novel “electronic nose” technology is highly intriguing, but [17] DiGiulio DB. Diversity of microbes in amniotic fluid. Semin Fetal Neonatal Med
studies that critically evaluate its predictive and diagnostic potential 2012;17:2–11.
remain to be done. [18] Aagaard K, Ma J, Antony KM, Ganu R, Petrosino J, Versalovic J. The placenta
harbors a unique microbiome. Sci Transl Med 2014;6. 237ra65.
The fact that certain microbial patterns as well as single microbes [19] Heida FH, van Zoonen A, Hulscher JBF, et al. A necrotizing enterocolitis-associated
are being found to be highly associated with NEC is intriguing. gut microbiota is present in the meconium: results of a prospective study. Clin Infect
However, the finding of an association is not adequate to directly base Dis 2016;62:863–70.
[20] Gosalbes MJ, Llop S, Valles Y, Moya A, Ballester F, Francino MP. Meconium mi-
therapeutic interventions. More rigorous studies of the causality will crobiota types dominated by lactic acid or enteric bacteria are differentially asso-
need to be done. These will need to be based on in-vitro as well as in- ciated with maternal eczema and respiratory problems in infants. Clin Exp Allergy
vivo techniques to recapitulate the disease in cell culture, and/or an- 2013;43:198–211.
[21] La Rosa PS, Warner BB, Zhou Y, et al. Patterned progression of bacterial populations
imal models. Once these are established, directed therapies using highly in the premature infant gut. Proc Natl Acad Sci USA 2014;111:12522–7.
specific antibiotics, bacteriophages, probiotics or other microbial ther- [22] Claud EC, Walker WA. Hypothesis: inappropriate colonization of the premature
apeutic techniques may be further evaluated for safety and efficacy. intestine can cause neonatal necrotizing enterocolitis. Faseb J 2001;15:1398–403.
[23] Claud EC, Walker WA. Bacterial colonization, probiotics, and necrotizing en-
Several studies suggest that a microbial therapeutic approach to terocolitis. J Clin Gastroenterol 2008;42(Suppl 2):S46–52.
NEC may be useful in the prevention of NEC. However, the most rig- [24] Nanthakumar NN, Fusunyan RD, Sanderson I, Walker WA. Inflammation in the
orous studies using probiotics to date are not supportive even though developing human intestine: a possible pathophysiologic contribution to necro-
tizing enterocolitis. Proc Natl Acad Sci USA 2000;97:6043–8.
numerous meta-analyses suggest efficacy. The fact that numerous dif-
[25] Nanthakumar N, Meng D, Goldstein AM, et al. The mechanism of excessive in-
ferent probiotics, which are very different in their character, have been testinal inflammation in necrotizing enterocolitis: an immature innate immune re-
utilized in these meta-analyses has been subject of the criticism. sponse. PLoS One 2011;6:e17776.
404
[26] Afrazi A, Sodhi CP, Richardson W, et al. New insights into the pathogenesis and pathogenesis of necrotizing enterocolitis. Semin Pediatr Surg 2005;14. 137–4.
treatment of necrotizing enterocolitis: Toll-like receptors and beyond. Pediatr Res [57] Denning PW, Maheshwari A. Necrotizing enterocolitis: hope on the horizon. Clin
2011;69:183–8. Perinatol 2013;40. xvii–xix.
[27] Morowitz MJ, Poroyko V, Caplan M, Alverdy J, Liu DC. Redefining the role of in- [58] Caplan MS, Simon D, Jilling T. The role of PAF, TLR, and the inflammatory response
testinal microbes in the pathogenesis of necrotizing enterocolitis. Pediatrics in neonatal necrotizing enterocolitis. Semin Pediatr Surg 2005;14:145–51.
2010;125:777–85. [59] Maheshwari A, Schelonka RL, Dimmitt RA, et al. Cytokines associated with ne-
[28] Brook I. Microbiology and management of neonatal necrotizing enterocolitis. Am J crotizing enterocolitis in extremely-low-birth-weight infants. Pediatr Res
Perinatol 2008;25:111–8. 2014;76:100–8.
[29] Balamurugan R, Magne F, Balakrishnan D, et al. Faecal bifidobacteria in Indian [60] Lu P, Sodhi CP, Hackam DJ. Toll-like receptor regulation of intestinal development
neonates; the effect of asymptomatic rotavirus infection during the first month of and inflammation in the pathogenesis of necrotizing enterocolitis. Pathophysiology
life. Indian J Med Res 2010;132:721–7. 2014;21:81–93.
[30] Mancini N, Poloniato A, Ghidoli N, et al. Potential role of the detection of en- [61] Sodhi C, Richardson W, Gribar S, Hackam DJ. The development of animal models
terobacterial DNA in blood for the management of neonatal necrotizing en- for the study of necrotizing enterocolitis. Dis Models Mech 2008;1:94–8.
terocolitis. J Med Microbiol 2012;61:1465–72. [62] Sodhi CP, Neal MD, Siggers R, et al. Intestinal epithelial Toll-like receptor 4 reg-
[31] Carlisle EM, Morowitz MJ. The intestinal microbiome and necrotizing enterocolitis. ulates goblet cell development and is required for necrotizing enterocolitis in mice.
Curr Opin Pediatr 2013;25:382–7. Gastroenterology 2012;143. 708–18 e1–5.
[32] Wang Y, Hoenig JD, Malin KJ, et al. 16S rRNA gene-based analysis of fecal mi- [63] Sodhi CP, Shi XH, Richardson WM, et al. Toll-like receptor-4 inhibits enterocyte
crobiota from preterm infants with and without necrotizing enterocolitis. ISME J proliferation via impaired beta-catenin signaling in necrotizing enterocolitis.
2009;3:944–54. Gastroenterology 2010;138:185–96.
[33] Musemeche CA, Kosloske AM, Bartow SA, Umland ET. Comparative effects of [64] Jilling T, Simon D, Lu J, et al. The roles of bacteria and TLR4 in rat and murine
ischemia, bacteria, and substrate on the pathogenesis of intestinal necrosis. models of necrotizing enterocolitis. J Immunol 2006;177:3273–82.
J Pediatr Surg 1986;21:536–8. [65] Smith PM, Howitt MR, Panikov N, et al. The microbial metabolites, short-chain fatty
[34] Cotten CM, Taylor S, Stoll B, et al. Prolonged duration of initial empirical antibiotic acids, regulate colonic Treg Cell Homeostasis. Science 2013;341:569–73.
treatment is associated with increased rates of necrotizing enterocolitis and death [66] Deshmukh HS, Liu Y, Menkiti OR, et al. The microbiota regulates neutrophil
for extremely low birth weight infants. Pediatrics 2009;123:58–66. homeostasis and host resistance to Escherichia coli K1 sepsis in neonatal mice. Nat
[35] Alexander VN, Northrup V, Bizzarro MJ. Antibiotic exposure in the newborn in- Med 2014;20:524–30.
tensive care unit and the risk of necrotizing enterocolitis. J Pediatr 2011;159:392–7. [67] Kabat AM, Srinivasan N, Maloy KJ. Modulation of immune development and
[36] Mai V, Torrazza RM, Ukhanova M, et al. Distortions in development of intestinal function by intestinal microbiota. Trends Immunol 2014;35:507–17.
microbiota associated with late onset sepsis in preterm infants. PLoS One [68] Arpaia N, Campbell C, Fan X, et al. Metabolites produced by commensal bacteria
2013;8:e52876. promote peripheral regulatory T-cell generation. Nature 2013;504:451–5.
[37] Warner BB, Deych E, Zhou Y, et al. Gut bacteria dysbiosis and necrotising en- [69] Hoverstad T, Midtvedt T. Short-chain fatty acids in germfree mice and rats. J Nutr
terocolitis in very low birthweight infants: a prospective case–control study. Lancet 1986;116:1772–6.
2016;387:1928–36. [70] Peng L, Li ZR, Green RS, Holzman IR, Lin J. Butyrate enhances the intestinal barrier
[38] Tyson GW, Chapman J, Hugenholtz P, et al. Community structure and metabolism by facilitating tight junction assembly via activation of AMP-activated protein ki-
through reconstruction of microbial genomes from the environment. Nature nase in Caco-2 cell monolayers. J Nutr 2009;139:1619–25.
2004;428:37–43. [71] Furusawa Y, Obata Y, Fukuda S, et al. Commensal microbe-derived butyrate induces
[39] Gill SR, Pop M, Deboy RT, et al. Metagenomic analysis of the human distal gut the differentiation of colonic regulatory T cells. Nature 2013;504:446–50.
microbiome. Science 2006;312:1355–9. [72] Canani RB, Costanzo MD, Leone L, Pedata M, Meli R, Calignano A. Potential ben-
[40] Relman DA. Microbial genomics and infectious diseases. N Engl J Med eficial effects of butyrate in intestinal and extraintestinal diseases. Wld J
2011;365:347–57. Gastroenterol 2011;17:1519–28.
[41] Raveh-Sadka T, Thomas BC, Singh A, et al. Gut bacteria are rarely shared by co- [73] Angrisano T, Pero R, Peluso S, et al. LPS-induced IL-8 activation in human intestinal
hospitalized premature infants, regardless of necrotizing enterocolitis development. epithelial cells is accompanied by specific histone H3 acetylation and methylation
eLife 2015;4. changes. BMC Microbiol 2010;10:172.
[42] Morowitz MJ, Denef VJ, Costello EK, et al. Strain-resolved community genomic [74] Oliver SG, Winson MK, Kell DB, Baganz F. Systematic functional analysis of the
analysis of gut microbial colonization in a premature infant. Proc Natl Acad Sci USA yeast genome. Trends Biotechnol 1998;16:373–8.
2011;108:1128–33. [75] Tuohy KM, Gougoulias C, Shen Q, Walton G, Fava F, Ramnani P. Studying the
[43] Ward DV, Scholz M, Zolfo M, et al. Metagenomic sequencing with strain-level re- human gut microbiota in the trans-omics era – focus on metagenomics and meta-
solution implicates uropathogenic E. coli in necrotizing enterocolitis and mortality bonomics. Curr Pharmaceut Des 2009;15:1415–27.
in preterm infants. Cell Rep 2016;14:2912–24. [76] Swann JR, Tuohy KM, Lindfors P, et al. Variation in antibiotic-induced microbial
[44] Torrazza RM, Neu J. The altered gut microbiome and necrotizing enterocolitis. Clin recolonization impacts on the host metabolic phenotypes of rats. J Proteome Res
Perinatol 2013;40:93–108. 2011;10:3590–603.
[45] Wojkowska-Mach J, Rozanska A, Borszewska-Kornacka M, et al. Necrotising en- [77] Ponnusamy K, Choi JN, Kim J, Lee SY, Lee CH. Microbial community and meta-
terocolitis in preterm infants: epidemiology and antibiotic consumption in the bolomic comparison of irritable bowel syndrome faeces. J Med Microbiol
Polish neonatology network neonatal intensive care units in 2009. PLoS One 2011;60:817–27.
2014;9:e92865. [78] Dai ZL, Zhang J, Wu G, Zhu WY. Utilization of amino acids by bacteria from the pig
[46] Cotten CM. Adverse consequences of neonatal antibiotic exposure. Curr Opin small intestine. Amino Acids 2010;39:1201–15.
Pediatr 2016;28:141–9. [79] Cevallos-Cevallos JM, Danyluk MD, Reyes-De-Corcuera JI. GC-MS based metabo-
[47] Torrazza RM, Ukhanova M, Wang X, et al. Intestinal microbial ecology and en- lomics for rapid simultaneous detection of Escherichia coli O157:H7, Salmonella
vironmental factors affecting necrotizing enterocolitis. PLoS One 2013;8:e83304. Typhimurium, Salmonella Muenchen, and Salmonella Hartford in ground beef and
[48] Terrin G, Passariello A, De Curtis M, et al. Ranitidine is associated with infections, chicken. J Food Sci 2011;76:M238–46.
necrotizing enterocolitis, and fatal outcome in newborns. Pediatrics [80] Stewart CJ, Nelson A, Treumann A, et al. Metabolomic and proteomic analysis of
2012;129:e40–5. serum from preterm infants with necrotising entercolitis and late-onset sepsis.
[49] Gupta RW, Tran L, Norori J, et al. Histamine-2 receptor blockers alter the fecal Pediatr Res 2016;79:425–31.
microbiota in premature infants. J Pediatr Gastroenterol Nutr 2013;56:397–400. [81] Morrow AL, Lagomarcino AJ, Schibler KR, et al. Early microbial and metabolomic
[50] Mehall JR, Northrop R, Saltzman DA, Jackson RJ, Smith SD. Acidification of for- signatures predict later onset of necrotizing enterocolitis in preterm infants.
mula reduces bacterial translocation and gut colonization in a neonatal rabbit Microbiome 2013;1:13.
model. J Pediatr Surg 2001;36:56–62. [82] Wilcock A, Begley P, Stevens A, Whatmore A, Victor S. The metabolomics of ne-
[51] Boneti C, Habib CM, Keller JE, et al. Probiotic acidified formula in an animal model crotising enterocolitis in preterm babies: an exploratory study. J Matern Fetal
reduces pulmonary and gastric bacterial load. J Pediatr Surg 2009;44:530–3. Neonatal Med 2016;29:758–62.
[52] Carrion V, Egan EA. Prevention of neonatal necrotizing enterocolitis. J Pediatr [83] De Magistris A, Corbu S, Cesare Marincola F, et al. NMR-based metabolomics
Gastroenterol Nutr 1990;11:317–23. analysis of urinary changes in neonatal necrotizing enterocolitis. 11th International
[53] Manzoni P, Garcia Sanchez R, Meyer M, et al. Exposure to gastric acid inhibitors Workshop on Neonatology; Cagliari, Italy, October 26th–31st, 2015. Abstract 34, p.
increases the risk of infection in preterm very low birth weight infants but con- 37–38.
comitant administration of lactoferrin counteracts this effect. J Pediatr 2018;193. [84] Stewart CJ, Embleton ND, Marrs EC, et al. Temporal bacterial and metabolic de-
62–7 e1. velopment of the preterm gut reveals specific signatures in health and disease.
[54] Romaine A, Ye D, Ao Z, et al. Safety of histamine-2 receptor blockers in hospitalized Microbiome 2016;4:67.
VLBW infants. Early Hum Dev 2016;99:27–30. [85] Embleton ND, Berrington JE, Dorling J, et al. Mechanisms affecting the gut of
[55] Hunt KM, Foster JA, Forney LJ, et al. Characterization of the diversity and temporal preterm infants in enteral feeding trials. Front Nutr 2017;4:14.
stability of bacterial communities in human milk. PLoS One 2011;6:e21313. [86] Cacho NT, Harrison NA, Parker LA, et al. Personalization of the microbiota of donor
[56] Neu J, Chen M, Beierle E. Intestinal innate immunity: how does it relate to the human milk with mother's own milk. Front Microbiol 2017;8:1470.
405

Enteral feeding composition and necrotizing enterocolitis T

∗
Diomel de la Cruz , Catalina Bazacliu
University of Florida, Department of Pediatrics, Division of Neonatology, Gainesville, FL, USA
Keywords: Enteral feeding and composition play a chief role in the prevention and treatment of necrotizing enterocolitis
Necrotizing enterocolitis (NEC). In the face of decades of research on this fatal disease, the exact mechanism of disease is still poorly
Human milk understood. There is established evidence that providing mother's own breast milk and standardization of
Donor breast milk feeding regimens leads to a decreased risk for NEC. More recent studies have focused on the provision of donor
Lactoferrin
human milk or an exclusive human milk diet in the endeavor to prevent NEC while still maintaining adequate
Human milk oligosaccharides
nutrition to the premature infant. There is growing literature on the provision of specific human milk compo-
nents and its effect on the incidence of NEC.
1. Introduction nutrition [12]. The benefits of human milk are attributed to the pre-
sence of bioactive substances that act with bactericidal and im-
Necrotizing enterocolitis (NEC) is the most common, life-threa- munomodulating properties. Evidence shows that exclusive human
tening gastrointestinal emergency in the neonatal period affecting ap- milk appears to decrease the risk of late-onset sepsis and feeding in-
proximately 7% of premature very low birth weight (VLBW) infants. tolerance and other complications of prematurity such as retinopathy of
Mortality rates range from 15% to 30%, depending on the need for prematurity [13,14]. Furthermore, provision of human milk to pre-
surgical intervention or disease severity [1,2]. NEC significantly in- mature infants is associated with improved neurodevelopmental out-
creases medical care costs, especially if surgery is required [3]. Total comes, shorter NICU stay, decreased hospital readmissions and lower
hospital costs for NEC are estimated at US$7 million per year [4,5]. The obesity and insulin resistance in adolescence [15–18].
death toll and the long-term sequelae make prevention and manage-
ment of this disease very important. Despite being studied for more 2.1. Exclusive human milk
than six decades, the pathophysiology of NEC is still incompletely
deciphered and poorly understood. The underlying cause is considered A large amount of data supports the concept that exclusive human
multifactorial, an association of intestinal and immune system im- milk intake may reduce the incidence and severity of NEC [19–22].
maturity, intestinal mucosal injury, bacterial dysbiosis, and type of Lucas et al. [23] demonstrated, in a landmark paper in 1990, that
feeds, being widely accepted [6–11]. We know that a vast majority of human breast milk has protective effects against NEC when compared
cases of NEC develop after the initiation of enteral feeding. Despite the with formula. Abrams et al. [24] performed a meta-analysis and re-
persistent challenges in the diagnosis and prevention of NEC, there has ported a lower risk of NEC in those who received an exclusive human
been considerable evidence for the past few decades that two feeding milk diet.
interventions decrease its incidence: provision of breastmilk and stan- One challenge that confronts an exclusive human milk diet is the
dardized feeding regimens. Specific human milk fractions and compo- matter of fortification. Human milk does not contain sufficient amounts
nents seemed promising for prevention of NEC in preclinical studies, of protein, calcium, phosphorous and calories for the ideal growth of
but there is scarce clinical evidence of significant impact on disease the premature infant [12]. Most neonatal intensive care units use bo-
prevention and treatment for any single one. vine-based fortifier that may promote an environment conducive to
NEC. The most recent meta-analysis looking at 14 trials on bovine
2. Human milk and necrotizing enterocolitis fortification of breastmilk reports that human breastmilk fortification
results in slightly improved growth and alkaline phosphatase, without
Human milk is established to be the most favorable form of infant increasing the risk of NEC [25]. Further studies on human milk-based
∗
Corresponding author. University of Florida, Department of Pediatrics, Division of Neonatology, 1600 SW Archer Road, PO Box 100296, Gainesville, FL 32610-
0296, USA.
E-mail address: ddelacruz@ufl.edu (D.d.l. Cruz).
D.d.l. Cruz, C. Bazacliu Seminars in Fetal and Neonatal Medicine 23 (2018) 406–410
human milk fortifier are still needed. or even a more aggressive daily advancement of enteral feeds
(> 20 mL/kg/day) [39,40].
2.2. Donor human milk
3.2. Early fortification of feeds
Questions remain regarding whether donor human milk has the
same protective effect as mother's own milk. The evidence is less per- It is generally recognized that fortification of human milk is re-
suasive, given that most studies do not randomize infants to purely quired to deliver adequate calories, calcium and phosphorous and
donor breast milk versus formula. More recent meta-analyses have protein to the preterm infant [25,44]. However, early fortification and
concluded that infants fed donor breastmilk had poor growth but less increasing osmolality of feeds has raised questions on its safety [45].
NEC compared to those fed formula [26,27]. Notably, in both meta- Two RCTs have shown that early fortification of enteral feeds did not
analyses, the benefit of preventing NEC was lost when donor human increase the risk of NEC or feeding intolerance while presenting the
milk was used to supplement mother's own milk. Furthermore, most of benefit of improved growth and protein delivery [46,47].
the studies used in the Cochrane review used studies published in the
1980s and excluded infants at the highest risk, namely the smallest 4. Colostrum
infants. Two randomized controlled trials (RCTs) have failed to de-
monstrate a reduction in NEC, whether on donor human milk or for- Produced in the first few days postpartum, colostrum has a higher
mula, when mother's own milk was unavailable [21,28]. We found no concentration of proteins and immunoactive components than the
clinical trial directly comparing the impact of donor milk with mother's mature milk [48]. The application of both bovine and human colostrum
own milk on the development of NEC. Given the current knowledge, an had been studied recently.
RCT may even be considered unethical. Recently, the DoMINO trial Studies in a preterm piglet NEC model suggest that bovine colos-
demonstrated no cognitive benefits when infants were randomized to trum protects against NEC improves the intestinal architecture and
donor breastmilk versus formula [29]. Donor human milk differs from decreases inflammation [49]. Based on these data, a feasibility and
mother's own breast milk with respect to caloric content and macro- tolerability pilot safety study on bovine colostrum administration in
nutrient composition, as well as the amount of bioactive ingredients. preterm infants was initiated in Denmark and China. The researchers
Holder pasteurization, performed to decrease bacteria and viruses, also found no difference in feeding tolerance or clinical adverse outcomes
diminishes or inactivates protective elements of mother's own milk, during the first two weeks of life, except a transient elevation of plasma
including secretory IgA, lactoferrin, commensal bacteria, and lysozymes tyrosine on day 7 in infants receiving bovine colostrum, an effect at-
[30,31]. tributed to their increased protein intake [50].
Several studies on the effect of oral human colostrum administration
2.3. Dosing of human milk in preterm infants and its relationship with the immune system con-
cluded with contradictory results. In a retrospective cohort, Seigel et al.
Multiple studies examined the effect of different doses of human demonstrated that colostrum could be safely administered orophar-
milk on the incidence of NEC. In the secondary analysis of a meta- yngeally to ELBW infants in the first two days of life and might provide
analysis, Abrams et al. [24] reported that the percentage of diet other the benefit of a better weight gain at 36 weeks [51]. This study was
than exclusive human milk was a significant predictor of NEC. The risk followed by an RCT of oral colostrum priming that showed significant
for NEC increased by 11–21% for every 10% increase in non-human reduction in the median length of hospitalization as compared to in-
milk in the diet. Two RCTs [21,32] reported that the incidence of NEC fants who did not receive oral colostrum, but no impact on salivary
appeared to decrease in those infants who received more than 50% of immune peptides and oral microbiota in preterm infants receiving oral
their total feeds made up of mother's own milk. The basis of this see- colostrum versus the control group [52]. Two other RCTs in preterm
mingly dose-dependent response is unclear and may be related to more infants demonstrated increase in salivary secretory IgA [53] and cir-
protection provided by a greater intake of human milk, or less formula. culating secretory immunoglobulin A [54]. Lee et al. noted a significant
decrease in clinical sepsis, as well as an increase in immunoprotective
3. Standardized feeding regimens and necrotizing enterocolitis factors and a decrease in pro-inflammatory cytokines in the colostrum
group [54].
The preterm gut is especially at risk for feeding intolerance. This is
most likely due to an immature gut barrier, microbial dysbiosis, limited 5. Human milk components
intestinal function and motility. A linear trajectory of feeding ad-
vancement is believed to be the most intuitive. Several studies have 5.1. Lactoferrin
shown that standardized feeding regimens result in a decreased in-
cidence of NEC [33–35]. Known for its antimicrobial properties, lactoferrin is the most
abundant protein in human milk whey [48] and may be effective in
3.1. Rate of advancement of feeds preventing NEC. Bovine lactoferrin (BLF) has a 69% similarity with the
human lactoferrin [48] but has a much lower concentration in bovine
At the present time, the subject of trophic feeding has divided milk. Animal studies demonstrated that lactoferrin might prevent in-
clinicians into those who delay the initiation of nutritive feeding, ex- testinal inflammation and NEC by interfering with activation of the pro-
tended the duration of trophic feeding for a few days to a week, and inflammatory cascade. It may compete with LPS in binding on Toll-like
those who advance enteral feedings without offering trophic feeding. receptor-4, inhibiting the activation of nuclear factor-κβ-regulated
There are clinical data suggesting that trophic feeding for a few days genes [55].
lead to a decreased risk of NEC when compared to early progressive A large multicenter trial comparing BLF, bovine lactoferrin with
feeding [36,37]. Conversely, prolonged trophic feeds have been shown Lactobacillus rhamnosus (BLF + LGG) and placebo in VLBW infants
to a delay time to full enteral feeding [38,39]. A delay in achievement showed a statistically significant reduction in NEC in the BLF + LGG
of full enteral feeds increases risk for nosocomial sepsis and infection, group and a trend towards decreasing NEC in BLF in comparison with
prolongs parenteral nutrition and central lines, extends length of hos- controls [56]. Of note, the incidence of NEC in the control group was
pital stay, and increases risk for postnatal growth restriction [38–43]. relatively low, at 5.3%. Another small trial from Turkey showed de-
Most current meta-analyses have not revealed any increased risk for creased sepsis but no difference in NEC, even though there was no NEC
NEC or feeding intolerance with early progression from trophic feeds, case in the lactoferrin-treated group [57]. A Cochrane review of
407
lactoferrin in preterm infants concluded that lactoferrin decreased NEC own milk combined with bovine milk fortifier, but the current evidence
stage II or greater by 30%, but the quality of evidence was marginal is insufficient to support this practice. Due to composition changes in-
[58]. Administration of recombinant human lactoferrin had no effect on duced by pasteurization and thawing to which human donor milk is
NEC rates, but significantly reduced in hospital-acquired infection in subjected, it does not seem to have the same protective effects against
the treatment group [59]. The results of the “Enteral lactoferrin in NEC without the presence of mother's own milk. Due to its undeniable
neonates” (ELFIN), a large RCT of lactoferrin in infants less than 32 benefits, extensive research aimed to identify protective components
weeks gestation that finished enrolling in 2017, is expected to shed has been done. Even though animal studies demonstrated that a variety
more light on the role lactoferrin may play in NEC prevention in the of human milk components seem to be effective in preventing and de-
future (www.npeu.ox.ac.uk). creasing severity of NEC, the few tested in human trials fail to show a
significant improvement. More research in this area is needed, though it
5.2. Human milk oligosaccharides faces a few problems. The animal models for NEC do not completely
recapitulate the premature infants' disease and it is likely that the array
Human milk oligosaccharides (HMOs) are a major bioactive com- of bioactive components from human milk act synergistically to
ponent of human milk that varies with maternal genetics and stage of dampen the immune system response and to nourish and protect the
lactation [60]. HMOs play an important role in microbial colonization intestinal epithelium. Clinically, NEC is a disease with an overall low
of the gut after birth, as suggested by the increase in fucosyltransferase incidence, making clinical trials require a high number of patients to
activity during the first postnatal weeks [61], and subsequently serve as enroll to adequately power for effect.
nutritional substrate for “good” bacteria influencing the development of
intestinal microbiome [60]. HMOs also inhibit bacterial pathogen ad-
herence to mucosal surfaces, protecting breast-fed infants against var- 6.1. Practice points
ious infections [62]. A sialated isomer, disialyllacto-N-tetraose, that
confers protection from NEC was identified in animal models by Jant- • Two feeding interventions decrease the risk for necrotizing en-
scher-Krenn et al. [63]. Sialylated galacto-oligosaccharides and 2′-fu- terocolitis: provision of mother's breast milk and standardization of
cosyllactose were also shown to reduce NEC in neonatal rodents [64]. feeding regimens.
Mechanistically, 2′-fucosyllactose improves intestinal perfusion by up- • Early progressive feeding and fortification are safe with the benefit
regulating endothelial nitric oxide synthase in a mouse model [65]. of better growth and shorter time to full feeds.
Colostrum HMOs can also be protective by attenuating the in- • Donor human breastmilk and human milk-based human milk for-
flammatory response and promoting the maturation of the intestinal tifier may contribute to NEC prevention.
mucosal immune system in ex-vivo immature intestinal cells culture • A higher dose of human breast milk (> 50% of feeds) reduces the
[66]. A large amount of fucosylated oligosaccharides, including se- incidence of NEC.
cretor antigen (H) are found in milk. The secretory status of the mother • Human milk components could not be recommended for prevention
determines the fucosylation of glycans in the milk [60] which may or treatment of NEC at this time.
further increase the risk of developing NEC in low or non-secretory
infants, as Morrow et al. demonstrated that premature infants with
6.2. Research directions
fucosyltransferase-2 low secretor phenotype have a ten-fold increase in
NEC [67].
• The role of human milk-based human milk fortification in the nu-
trition of the preterm infant and the prevention of NEC.
5.3. Other milk components
• The role of donor human milk in the nutrition of the extremely
premature infant.
A few other human milk components that have been studied for
their potential in preventing NEC are listed in Table 1.
• The role of oral colostrum in modulation of immune system and
microbiota.
6. Summary
• The role of individual or combination of human milk component in
prevention and treatment of disease.
Enteral feeds with mother's own milk and consistent feeding prac-
tices are the only proven and modifiable interventions that help prevent Conflicts of interest
a significant number of NEC cases. Exclusive human milk diet could,
theoretically, be better than the more customary combined mother's None declared.
Table 1
Milk components contributing to necrotizing enterocolitis (NEC) prevention.
Component Proposed mechanism Human studies
Transforming growth factor- • Bound to glycans: bioavailability is increased by • Highest in colostrum, decreased during lactation [69]
β2 low pH [48] • small
Low TGF-β2 in the milk of mothers whose infants developed NEC, but the study had a
• Decreases
[68]
inflammation and promotes repair sample size and was not powered for NEC [69]
Immunoglobulins • Trials used IgG, IgG/IgA and IgG/IgA/IgM

• Cochrane review of oral immunoglobulins for NEC prophylaxis concluded that oral Ig
did not result in significant reduction of NEC [70]
• No trial for IgA alone was done
Glutamine • Decreases gut permeability and preserves villi
structure in mature human intestine [71]
• Cochrane review meta-analysis of glutamine supplementation oral or enteral to
prevent mortality and morbidity included 12 studies did not show any differences in
mortality or morbidities, including NEC [72]
Arginine • Plasma concentration lower in infants who • Three small trials with contradictory results
developed NEC [73] • Cochrane review found some reduction in NEC stage I and III, but concluded that there
is insufficient evidence to support supplementation [74]
408
Funding sources very low birth weight infants during the first 60 days of life. Neonatology
2012;102:276–81.
[33] Wiedmeier S, Henry E, Baer V, et al. Center differences in NEC within one health-
None. care system may depend on feeding protocol. Am J Perinatol 2008;25:5–11.
[34] Patole SK, de Klerk N. Impact of standardised feeding regimens on incidence of
References neonatal necrotising enterocolitis: a systematic review and meta-analysis of ob-
servational studies. Arch Dis Child Fetal Neonatal Ed 2005;90:F147–51.
[35] Jasani B, Patole S. Standardized feeding regimen for reducing necrotizing en-
[1] Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011;364:255–64. terocolitis in preterm infants: an updated systematic review. J Perinatol
[2] Horbar JD, Carpenter JH, Badger GJ, et al. Mortality and neonatal morbidity among 2017;37:827–33.
infants 501 to 1500 grams from 2000 to 2009. Pediatrics 2012;129:1019–26. [36] Berseth CL, Bisquera JA, Paje VU. Prolonging small feeding volumes early in life
[3] Ganapathy V, Hay JW, Kim JH. Costs of necrotizing enterocolitis and cost-effec- decreases the incidence of necrotizing enterocolitis in very low birth weight infants.
tiveness of exclusively human milk-based products in feeding extremely premature Pediatrics 2003;111:529–34.
infants. Breastfeed Med 2012;7:29–37. [37] Henderson G, Craig S, Brocklehurst P, McGuire W. Enteral feeding regimens and
[4] Ganapathy V, Hay JW, Kim JH, Lee ML, Rechtman DJ. Long term healthcare costs of necrotising enterocolitis in preterm infants: a multicentre case–control study. Arch
infants who survived neonatal necrotizing enterocolitis: a retrospective longitudinal Dis Child Fetal Neonatal Ed 2009;94:F120–3.
study among infants enrolled in Texas Medicaid. BMC Pediatr 2013;13:127. [38] Leaf A, Dorling J, Kempley S, et al. Early or delayed enteral feeding for preterm
[5] Bisquera JA, Cooper TR, Berseth CL. Impact of necrotizing enterocolitis on length of growth-restricted infants: a randomized trial. Pediatrics 2012;129:e1260–8.
stay and hospital charges in very low birth weight infants. Pediatrics [39] Morgan J, Young L, Mcguire W. Delayed introduction of progressive enteral feeds to
2002;109:423–8. prevent necrotising enterocolitis in very low birth weight infants. Cochrane
[6] Neu J. Necrotizing enterocolitis: the mystery goes on. Neonatology Database Syst Rev 2014;12. CD001970.
2014;106:289–95. [40] Morgan J, Young L, Mcguire W. Slow advancement of enteral feed volumes to
[7] Torrazza RM, Neu J. The altered gut microbiome and necrotizing enterocolitis. Clin prevent necrotising enterocolitis in very low birth weight infants. Cochrane
Perinatol 2013;40:93–108. Database Syst Rev 2014;12. CD001241.
[8] Patole S. Prevention and treatment of necrotising enterocolitis in preterm neonates. [41] Stoll BJ, Hansen NI, Adams-Chapman I, et al. Neurodevelopmental and growth
Early Hum Dev 2007;83:635–42. impairment among extremely low-birth-weight infants with neonatal infection.
[9] Morowitz MJ, Poroyko V, Caplan M, Alverdy J, Liu DC. Redefining the role of in- J Am Med Assoc 2004;292:2357.
testinal microbes in the pathogenesis of necrotizing enterocolitis. Pediatrics [42] Karagol BS, Zenciroglu A, Okumus N, Polin RA. Randomized controlled trial of slow
2010;125:777–85. vs rapid enteral feeding advancements on the clinical outcomes of preterm infants
[10] Carlisle EM, Morowitz MJ. The intestinal microbiome and necrotizing enterocolitis. with birth weight 750–1250 g. J Parenter Enteral Nutr 2013;37:223–8.
Curr Opin Pediatr 2013;25:382–7. [43] Ehrenkranz RA, Dusick AM, Vohr BR, Wright LL, Wrage LA, Poole WK. Growth in
[11] Lin PW, Stoll BJ. Necrotising enterocolitis. Lancet 2006;368(9543):1271–83. the neonatal intensive care unit influences neurodevelopmental and growth out-
[12] Section on Breastfeeding. Breastfeeding and the use of human milk. Pediatrics comes of extremely low birth weight infants. Pediatrics 2006;117:1253–61.
2012;129:e827–41. [44] Picaud J-C, Houeto N, Buffin R, Loys C-M, Godbert I, Haÿs S. Additional protein
[13] Bharwani SK, Green BF, Pezzullo JC, Bharwani SS, Bharwani SS, Dhanireddy R. fortification is necessary in extremely low-birth-weight infants fed human milk.
Systematic review and meta-analysis of human milk intake and retinopathy of J Pediatr Gastroenterol Nutr 2016;63:103–5.
prematurity: a significant update. J Perinatol 2016;36:913–20. [45] Rosas R, Sanz M, Fernández-Calle P, et al. Experimental study showed that adding
[14] Patel AL, Johnson TJ, Engstrom JL, et al. Impact of early human milk on sepsis and fortifier and extra-hydrolysed proteins to preterm infant mothers' milk increased
health-care costs in very low birth weight infants. J Perinatol 2013;33:514–9. osmolality. Acta Paediatr 2016;105:e555–60.
[15] Lechner BE, Vohr BR. Neurodevelopmental outcomes of preterm infants fed human [46] Shah SD, Dereddy N, Jones TL, Dhanireddy R, Talati AJ. Early versus delayed
milk: a systematic review. Clin Perinatol 2017;44:69–83. human milk fortification in very low birth weight infants – a randomized controlled
[16] Singhal A, Cole TJ, Fewtrell M, Lucas A. Breastmilk feeding and lipoprotein profile trial. J Pediatr 2016;174. 126–31.e1.
in adolescents born preterm: follow-up of a prospective randomised study. Lancet [47] Mukhopadhyay K, Narnag A, Mahajan R. Effect of human milk fortification in ap-
2004;363(9421):1571–8. propriate for gestation and small for gestation preterm babies: a randomized con-
[17] Singhal A, Cole TJ, Lucas A. Early nutrition in preterm infants and later blood trolled trial. Indian Pediatr 2007;44:286–90.
pressure: two cohorts after randomised trials. Lancet 2001;357(9254):413–9. [48] Ballard O, Morrow AL. Human milk composition. Pediatr Clin 2013;60:49–74.
[18] Isaacs EB, Fischl BR, Quinn BT, Chong WK, Gadian DG, Lucas A. Impact of breast [49] Shen RL, Thymann T, Østergaard MV, et al. Early gradual feeding with bovine
milk on intelligence quotient, brain size, and white matter development. Pediatr colostrum improves gut function and NEC resistance relative to infant formula in
Res 2010;67:357–62. preterm pigs. Am J Physiol Gastrointest Liver Physiol 2015;309:G310–23.
[19] Sullivan S, Schanler RJ, Kim JH, et al. An exclusively human milk-based diet is [50] Juhl SM, Ye X, Zhou P, et al. Bovine colostrum for preterm infants in the first days of
associated with a lower rate of necrotizing enterocolitis than a diet of human milk life. J Pediatr Gastroenterol Nutr 2018;66:471–8.
and bovine milk-based products. J Pediatr 2010;156. 562–567.e1. [51] Seigel JK, Smith PB, Ashley PL, et al. Early administration of oropharyngeal co-
[20] Cristofalo EA, Schanler RJ, Blanco CL, et al. Randomized trial of exclusive human lostrum to extremely low birth weight infants. Breastfeed Med 2013;8:491–5.
milk versus preterm formula diets in extremely premature infants. J Pediatr [52] Romano-Keeler J, Azcarate-Peril MA, Weitkamp J-H, et al. Oral colostrum priming
2013;163. 1592–5.e1. shortens hospitalization without changing the immunomicrobial milieu. J Perinatol
[21] Schanler RJ, Lau C, Hurst NM, Smith EO. Randomized trial of donor human milk 2017;37:36–41.
versus preterm formula as substitutes for mothers' own milk in the feeding of ex- [53] Glass K, Greecher C, Doheny K. Oropharyngeal administration of colostrum in-
tremely premature infants. Pediatrics 2005;116:400–6. creases salivary secretory IgA levels in very low-birth-weight infants. Am J
[22] Meinzen-Derr J, Poindexter B, Wrage L, Morrow AL, Stoll B, Donovan EF. Role of Perinatol 2017;34:1389–95.
human milk in extremely low birth weight infants' risk of necrotizing enterocolitis [54] Lee J, Kim H-S, Jung YH, et al. Oropharyngeal colostrum administration in ex-
or death. J Perinatol 2009;29:57–62. tremely premature infants: an RCT. Pediatrics 2015;135:e357–66.
[23] Lucas A, Cole TJ. Breast milk and neonatal necrotising enterocolitis. Lancet [55] Actor JK, Hwang S-A, Kruzel ML. Lactoferrin as a natural immune modulator. Curr
336(8730):1519–1523. Pharmaceut Des 2009;15:1956–73.
[24] Abrams SA, Schanler RJ, Lee ML, Rechtman DJ. Greater mortality and morbidity in [56] Manzoni P, Meyer M, Stolfi I, et al. Bovine lactoferrin supplementation for pre-
extremely preterm infants fed a diet containing cow milk protein products. vention of necrotizing enterocolitis in very-low-birth-weight neonates: a rando-
Breastfeed Med 2014;9:281–5. mized clinical trial. Early Hum Dev 2014;90:S60–5.
[25] Brown JV, Embleton ND, Harding JE, McGuire W. Multi-nutrient fortification of [57] Akin IM, Atasay B, Dogu F, et al. Oral lactoferrin to prevent nosocomial sepsis and
human milk for preterm infants. Cochrane Database Syst Rev 2016;5. CD000343. necrotizing enterocolitis of premature neonates and effect on T-regulatory cells. Am
[26] Quigley M, McGuire W. Formula versus donor breast milk for feeding preterm or J Perinatol 2014;31:1111–20.
low birth weight infants. Cochrane Database Syst Rev 2014;4. CD002971. [58] Pammi M, Abrams SA. Oral lactoferrin for the prevention of sepsis and necrotizing
[27] Boyd CA, Quigley MA, Brocklehurst P. Donor breast milk versus infant formula for enterocolitis in preterm infants. Cochrane Database Syst Rev 2015;2. CD007137.
preterm infants: systematic review and meta-analysis. Arch Dis Child Fetal Neonatal [59] Sherman MP, Adamkin DH, Niklas V, et al. Randomized controlled trial of ta-
Ed 2007;92:F169–75. lactoferrin oral solution in preterm infants. J Pediatr 2016;175. 68–73.e3.
[28] Corpeleijn WE, de Waard M, Christmann V, et al. Effect of donor milk on severe [60] Underwood MA, Gaerlan S, De Leoz MLA, et al. Human milk oligosaccharides in
infections and mortality in very low-birth-weight infants. JAMA Pediatr premature infants: absorption, excretion and influence on the intestinal microbiota.
2016;170:654. Pediatr Res 2015;78:670–7.
[29] O'Connor DL, Gibbins S, Kiss A, et al. Effect of supplemental donor human milk [61] Nanthakumar NN, Dai D, Newburg DS, Walker WA. The role of indigenous mi-
compared with preterm formula on neurodevelopment of very low-birth-weight croflora in the development of murine intestinal fucosyl- and sialyltransferases.
infants at 18 months. J Am Med Assoc 2016;316:1897. Faseb J 2003;17:44–6.
[30] Peila C, Moro GE, Bertino E, et al. The effect of holder pasteurization on nutrients [62] Morrow AL, Ruiz-Palacios GM, Jiang X, Newburg DS. Human-milk glycans that
and biologically-active components in donor human milk: a review. Nutrients inhibit pathogen binding protect breast-feeding infants against infectious diarrhea.
2016;8(8). J Nutr 2005;135:1304–7.
[31] Ewaschuk JB, Unger S, O'Connor DL, et al. Effect of pasteurization on selected [63] Jantscher-Krenn E, Zherebtsov M, Nissan C, et al. The human milk oligosaccharide
immune components of donated human breast milk. J Perinatol 2011;31:593–8. disialyllacto-N-tetraose prevents necrotising enterocolitis in neonatal rats. Gut
[32] Corpeleijn WE, Kouwenhoven SMP, Paap MC, et al. Intake of own mother's milk 2012;61:1417–25.
during the first days of life is associated with decreased morbidity and mortality in [64] Autran CA, Schoterman MHC, Jantscher-Krenn E, Kamerling JP, Bode L. Sialylated
409
galacto-oligosaccharides and 2′-fucosyllactose reduce necrotising enterocolitis in [69] Frost BL, Jilling T, Lapin B, Maheshwari A, Caplan MS. Maternal breast milk
neonatal rats. Br J Nutr 2016;116:294–9. transforming growth factor-beta and feeding intolerance in preterm infants. Pediatr
[65] Good M, Sodhi CP, Yamaguchi Y, et al. The human milk oligosaccharide 2′-fuco- Res 2014;76:386–93.
syllactose attenuates the severity of experimental necrotising enterocolitis by en- [70] Foster JP, Seth R, Cole MJ. Oral immunoglobulin for preventing necrotizing en-
hancing mesenteric perfusion in the neonatal intestine. Br J Nutr terocolitis in preterm and low birth weight neonates. Cochrane Database Syst Rev
2016;116:1175–87. 2016;4. CD001816.
[66] He Y, Liu S, Leone S, Newburg DS. Human colostrum oligosaccharides modulate [71] van der Hulst RR, van Kreel BK, von Meyenfeldt MF, et al. Glutamine and the
major immunologic pathways of immature human intestine. Mucosal Immunol preservation of gut integrity. Lancet 1993;341(8857):1363–5.
2014;7:1326–39. [72] Moe-Byrne T, Brown JVE, McGuire W. Glutamine supplementation to prevent
[67] Morrow AL, Meinzen-Derr J, Huang P, et al. Fucosyltransferase 2 non-secretor and morbidity and mortality in preterm infants. Cochrane Database Syst Rev 2016;1.
low secretor status predicts severe outcomes in premature infants. J Pediatr CD001457.
2011;158:745–51. [73] Becker RM, Wu G, Galanko JA, et al. Reduced serum amino acid concentrations in
[68] Maheshwari A, Kelly DR, Nicola T, et al. TGF-β2 suppresses macrophage cytokine infants with necrotizing enterocolitis. J Pediatr 2000;137:785–93.
production and mucosal inflammatory responses in the developing intestine. [74] Shah PS, Shah VS, Kelly LE. Arginine supplementation for prevention of necrotising
Gastroenterology 2011;140:242–53. enterocolitis in preterm infants. Cochrane Database Syst Rev 2017;4. CD004339.
410

The cost of necrotizing enterocolitis in premature infants T

a,∗ b c,d
Meredith E. Mowitz , Dmitry Dukhovny , John A.F. Zupancic
a
Division of Neonatology, University of Florida, Gainesville, FL, USA
b
Department of Pediatrics, Oregon Health & Science University, Portland, OR, USA
c
Department of Neonatology, Beth Israel Deaconess Medical Center, Boston, MA, USA
d
Department of Pediatrics, Harvard Medical School, Boston, MA, USA
Keywords: Necrotizing enterocolitis (NEC), a common morbidity of prematurity, affects 5–10% of premature infants with a
Necrotizing enterocolitis birthweight < 1500 g. The added cost remains unclear. Multiple studies report the cost of care for an infant with
Costs NEC as higher than that of well premature infants, but these studies are fraught with limitations. Surgical
Premature infant intervention and type of surgery appear to impact overall costs. Health care resource utilization extends beyond
the birth hospitalization, particularly in those infants requiring surgery, and persists to at least three years of age.
This narrative review of the literature reveals a paucity of studies and significant methodological deficiencies in
most included studies. Further studies of the cost of NEC need to address the issues of significant confounding in
this complex population.
1. Introduction 2. Overall cost and length of stay for birth hospitalization
Although enterocolitis-like disorders are seen in other populations, Although premature infants account for only a small percentage of
classically defined, necrotizing enterocolitis (NEC) is an inflammatory newborn admissions, they account for a disproportionately large per-
disorder of the intestines in premature infants. Similar disorders such as centage of costs, with a reported $5.8 billion spent on hospital care of
those seen in term infants and early spontaneous intestinal perforations preterm infants in the USA in 2001 [5], with even higher estimates of
are thought to involve differing pathogenesis and thus different ther- $26.2 billion from the societal perspective [6]. Furthermore, many
apeutic options. Despite advances in neonatal care, the incidence of studies suggest a negative correlation between costs and gestational age
NEC has remained stable, with classical NEC occurring in approxi- as well as a positive correlation with common morbidities [7]. Russell
mately 5–10% of very low birth weight (VLBW) infants (i.e. birth- et al. reported that mean hospital costs for preterm infants with
weight < 1500 g) [1]. NEC requires intensive medical care and high common morbidities of prematurity are four to seven times higher than
resource use, which may increase overall costs in the birth hospitali- their gestational age equivalent healthy controls [5]. Unfortunately, as
zation. Whereas NEC is not the most common morbidity of prematurity, NEC is a disease that primarily affects the preterm population, with
it often co-occurs with other common morbidities such as broncho- higher risks in the smallest and most premature infants, their additional
pulmonary dysplasia, brain injury and sepsis, making it difficult to resource utilization and costs of care due to this morbidity are com-
discern which disease process is increasing the overall costs [2]. Ad- pounded by their already higher baseline costs.
ditionally, the long-term complications and adverse neurodevelop- Many studies report the cost of care for infants with NEC. Although
mental outcomes associated with NEC may increase health care re- these studies take various methodological approaches they consistently
source utilization beyond the birth hospitalization [3]. Finally, show increased costs, especially compared to gestational age-matched
mortality with NEC remains high, with up to 42% mortality reported in infants without NEC and more so to healthy term infants.
the smallest premature infants [4]. This article outlines what is cur- Johnson et al. reported a mean direct cost of $100,752 (2009 US
rently known about the costs of birth hospitalization, surgical man- dollars) for VLBWs with NEC, a cost 1.4 times higher than in those
agement, long-term care and preventive strategies of NEC in the VLBW without NEC [8], whereas Russell et al. found that the median cost of
infant. birth hospitalization was $73,000 (2001 US dollars) for infants with
NEC compared to a median of $6800 in those without NEC, making it
∗
Corresponding author. Department of Pediatrics, Division of Neonatology, University of Florida, PO Box 100296, Gainesville, FL, 32610, USA.
E-mail address: mmowitz@peds.ufl.edu (M.E. Mowitz).
M.E. Mowitz et al. Seminars in Fetal and Neonatal Medicine 23 (2018) 416–419
the second most costly morbidity of prematurity [5,8]. The absolute peritoneal drain placement in a randomized trial, reporting some dif-
amount of money associated with a diagnosis of NEC is difficult to as- ferences, not reaching statistical significance, in secondary outcomes,
certain from these studies given methodologic differences and inflation, specifically in length of stay and days of total parenteral nutrition. This
but the increase in cost of care is consistent. Whereas this increase in lack of significance may be due the fact that the trial had insufficient
cost is multifactorial, some of the primary drivers identified include power to detect a difference; further the study did not specifically ad-
increased length of stay, increased use of parental nutrition, as well as dress the issue of cost or other specific resource utilization [15].
additional diagnostic and therapeutic interventions in infants with NEC
compared to those without NEC [8,9]. 4. Long-term costs
In the NEC population, length of stay is longer due to delayed
feeding, feeding intolerance, and need for medical and surgical inter- Following the initial NEC insult and recovery during the birth
ventions. Infants with NEC average a 16 day longer stay in comparison hospitalization, infants with NEC are at risk for feeding difficulties,
to those infants without NEC [8]. This difference is more apparent when failure to thrive, malabsorption, short-bowel syndrome, and develop-
separating out those infants requiring any surgical intervention for mental delay as they age [17,18]. These conditions are known to have
NEC. Bisquera et al. reported a mean length of stay of 142 days for increased health resource utilization beyond the neonatal intensive care
survivors of surgical NEC compared to 82 days in matched controls [9]. unit (NICU) stay in other conditions, but, unfortunately, few studies
Additionally, these infants were more likely to require central line have fully examined the costs of these longer-term morbidities for this
placement and have other comorbidities (e.g. intraventricular hemor- patient population. Ganapathy et al. reported long-term health resource
rhage, sepsis, and bronchopulmonary dysplasia) which may or may not utilization and cost in a cohort of Medicaid-enrolled Texas newborns.
be related to NEC but nonetheless compound the cost of care [8,9]. Those with medical NEC, compared to their matched controls, had
higher resource utilization and cost between 6 and 12 months of age but
3. Cost of medical and surgical interventions those differences did not persist beyond 12 months of age [17]. In the
6–12-month period, infants with medical NEC incurred an average of
The medical management of NEC typically includes holding of $5000 (2009 US dollars) more in costs than did matched controls [17].
enteral feeds with nutrition provided parenterally. This delay in feeding Infants with surgical NEC, on the other hand, had higher resource uti-
can increase length of stay, as difficulty in progression to oral feeds and lization and cost, which persisted to 36 months of age. Infants with
oral aversion may result. Furthermore, broad-spectrum antibiotics surgical NEC had mean overall health care costs of $45,213, $46,378
combined with close clinical observation with abdominal imaging are and $26,055 (2009 US dollars) at 6–12 months, 12–24 months and
also typically performed. Parenteral nutrition and antibiotics often lead 24–36 months of age, respectively [17]. In this group, costs of inpatient
to placement of a central line, which is associated with an increased risk care accounted for the majority of total costs from 6 to 12 months of
of central line-associated bacterial infections. These factors are just a age, with costs of home health care accounting for the majority of
few of the several plausible drivers of increased cost for patients with spending between 12 and 36 months [17]. Although this study begins
NEC. Unfortunately, these have been incompletely characterized in the to answer the question of long-term cost and resource utilization in
literature and their specific contribution to overall cost remains un- these patients, it fails to include the out-of-pocket costs, wage loss and
known. productivity loss that families and communities experience due to this
Surgical intervention for NEC occurs in approximately 40% of NEC disease. Few studies in neonatology have taken the societal perspective
cases [10]. To date, a superior strategy for surgical intervention has not with the inclusion of parental out-of-pocket expenses and wage loss,
been identified, therefore variation in surgical management of infants although this is advocated as an appropriate value-based care approach
with NEC persists [11,12]. Options for initial surgical intervention ty- [19–21].
pically include placement of a percutaneous peritoneal drain or ex-
ploratory laparotomy. Additionally, those undergoing initial peritoneal 5. Cost of preventive strategies
drain placement may proceed to laparotomy at a later date. For those
infants requiring surgical intervention, both length of stay and costs In the last decade much focus has been directed at preventive
have been reported higher than those requiring only medical manage- strategies for NEC. Although many strategies and therapies have been
ment [9,13]. Johnson et al. report surgery as a major contributor to the explored, a breast milk-based diet, in the form of mother's own milk,
cost of hospitalization in patients with NEC, finding, after adjustment donor milk and/or human milk-based fortification, continues to be the
for confounders, an increase in direct medical costs of $22,328 in pa- only measure with evidence of efficacy and safety [1]. Given the ex-
tients with surgical NEC versus $13,136 in patients with medical NEC, pense of these therapies, with donor human milk costs ranging from
although it is unclear what type(s) of surgery the infants underwent [8]. $4.00 to $5.00 and human milk-based fortifier costs of approximately
Stey et al. compared resource utilization and cost in infants with NEC $6.25/mL [22,23], the number needed to treat and cost (i.e. cost-ef-
based on their surgical intervention. The group undergoing peritoneal fectiveness) should be critically considered.
drain placement followed by laparotomy had higher length of stay Johnson et al. reported a decrease in hospitalization costs of $534
(median: 104 days) and cost (median: $318,259; 2013 US dollars) [14]. for every increase of 1 mL/kg/day of human milk received in the first
Those undergoing laparotomy alone had a median cost of $282,854 14 days of life [2]. Additional studies have examined the use of human
whereas those undergoing peritoneal drain placement alone had the donor milk only and with human milk-based fortifier to create an ex-
lowest median cost of $225,650 [14]. Although these studies point to clusive human milk diet. Ganapathy et al. created a model using results
differences in resource utilization and cost by surgical intervention, one of a randomized controlled trial of human milk-based fortifier and a
must take into account the lack of randomization for these studies single-state, hospital discharge database to compare the costs of care in
leading to potential residual confounding of why one approach is those receiving human milk-based fortifier versus bovine-based for-
chosen over another that cannot be adjusted away. Therefore, evalua- tifier. In this study, the costs of NEC outweighed the increased costs of
tion of these differences in resource utilization and cost by treatment fortification ($10,321 per infant receiving human milk-based fortifica-
approach may be considered as surgeons work toward a reference tion) with a net savings of $8167 per infant given human milk-based
standard in surgical intervention for NEC. To date, there is only one fortifier, although the results were sensitive to multiple parameters of
large randomized controlled trial published on surgical management by the model (e.g. NEC rates, costs) [24]. Arnold et al. created three
laparotomy versus peritoneal drain [15] with another large trial [16] modeling scenarios assuming that human milk reduced length of stay
due to complete in spring 2019. In the one reported trial, Moss et al. and NEC, reporting a cost saving with the use of donor milk in all
compared those undergoing primary laparotomy versus primary scenarios, although no sensitivity analyses were performed to test the
417
assumptions [25]. These models should be interpreted with caution and 7. Conclusion
placed in the smaller context of the individual NICUs where NEC rate
and prevention strategies vary. The cost of NEC is increased above the baseline for preterm infants
To date, only one formal economic evaluation alongside a rando- without NEC and further increased in those requiring surgical inter-
mized trial of donor milk has been undertaken [26]. The Donor Milk for vention. However, given the paucity of data, the degree of certainty
Improved Neurodevelopmental Outcomes (DoMINO) trial enrolled 363 around those estimates is clouded with confounding and likely over-
VLBW infants randomized to receive donor milk or preterm formula. shoots the incremental differences. It is difficult to separate out the
There was no difference in neurodevelopmental outcomes (primary costs specific to NEC in observational studies as these infants often have
outcome) but a decreased rate of NEC stage II or greater was found in multiple comorbidities that are confounding. To date, the literature
the donor milk group compared to the formula group (secondary out- exploring the costs of NEC is scarce and complicated by severe limita-
come) [27]. The economic evaluation of DoMINO used the outcomes of tions related to its observational nature and inability to fully address
NEC (stage I of greater, lack of feeding and a course of antibiotics for at confounding, therefore should be interpreted with caution. Further
least seven days) to compare costs between the groups receiving for- large, prospective, more methodologically appropriate studies of cost of
mula versus donor milk. Additionally, this study included parent illness in premature infants and specifically NEC are needed to better
questionnaires to evaluate out-of-pocket costs and wage loss and eval- understand the impact of novel preventive strategies.
uated outcomes and cost to 18 months [26]. No difference was found in
total costs over the trial period between the groups, although when Conflicts of interest
costs were separated into birth hospitalization versus post-discharge
time-frames, a significant difference was seen with the donor milk None declared.
group having lower post-discharge costs driven by less wage loss for the
parents [26]. This again emphasizes the importance of inclusion of costs Funding sources
beyond the third-party payer, as affirmed in the recently updated re-
commendations for cost-effectiveness analyses in health and medicine None.
[28]. The more inclusive approach provides a more comprehensive
perspective, accounts for the disease burden on the family as well as References
cost-shifting that may result from interventions that shorten lengths of
stay, for example. [1] Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011;364:255–64.
[2] Johnson TJ, Patel AL, Bigger HR, Engstrom JL, Meier PP. Cost savings of human
milk as a strategy to reduce the incidence of necrotizing enterocolitis in very low
birth weight infants. Neonatology 2015;107:271–6.
6. Future directions [3] Hintz SR, Kendrick DE, Stoll BJ, et al. Neurodevelopmental and growth outcomes of
extremely low birth weight infants after necrotizing enterocolitis. Pediatrics
2005;115:696–703.
Necrotizing enterocolitis is a disease of preterm infants associated [4] Fitzgibbons SC, Ching Y, Yu D, et al. Mortality of necrotizing enterocolitis expressed
with high morbidity and mortality. It is also associated with increased by birth weight categories. J Pediatr Surg 2009;44:1072–5. discussion 5–6.
costs that are difficult to determine due to the paucity of literature and [5] Russell RB, Green NS, Steiner CA, et al. Cost of hospitalization for preterm and low
birth weight infants in the United States. Pediatrics 2007;120:e1–9.
the substantial variation in the reported cost estimates. Observational [6] Institute of Medicine. Preterm birth: causes, consequences, and prevention.
studies on cost of illness are severely limited by its potential con- Washington DC: The National Academies Press; 2007https://doi.org/10.17226/
founding. In premature infants this is highlighted in that infants with 11622.
[7] Soilly AL, Lejeune C, Quantin C, Bejean S, Gouyon JB. Economic analysis of the
NEC typically have additional comorbidities which may or may not be
costs associated with prematurity from a literature review. Publ Hlth
directly related to NEC [8]. Observational studies to date have often 2014;128:43–62.
failed to adequately address these issues, using illness acuity adjust- [8] Johnson TJ, Patel AL, Jegier BJ, Engstrom JL, Meier PP. Cost of morbidities in very
ments other more advanced statistical methods such as propensity score low birth weight infants. J Pediatr 2013;162:243–9. e1.
[9] Bisquera JA, Cooper TR, Berseth CL. Impact of necrotizing enterocolitis on length of
matching. Randomized controlled trials present the best opportunity for stay and hospital charges in very low birth weight infants. Pediatrics
economic evaluations given the balance of known and unknown con- 2002;109:423–8.
founders from randomization. This approach, coupled with careful cost- [10] Guthrie SO, Gordon PV, Thomas V, Thorp JA, Peabody J, Clark RH. Necrotizing
enterocolitis among neonates in the United States. J Perinatol 2003;23:278–85.
accounting, would provide the least biased economic assessment be- [11] Rao SC, Basani L, Simmer K, Samnakay N, Deshpande G. Peritoneal drainage versus
tween the interventions to prevent NEC. Furthermore, considering the laparotomy as initial surgical treatment for perforated necrotizing enterocolitis or
high cost of more recent preventive strategies (e.g. donor milk, human spontaneous intestinal perforation in preterm low birth weight infants. Cochrane
Database Syst Rev 2011;6. CD006182.
milk fortifier, and probiotics), care must be taken to assess the cost- [12] Zani A, Eaton S, Puri P, et al. International survey on the management of necro-
benefit or return on investment. In the absence of randomized con- tizing enterocolitis. Eur J Pediatr Surg 2015;25:27–33.
trolled trials, a patient-level cost (or cost-effectiveness) analysis should [13] Abdullah F, Zhang Y, Camp M, et al. Necrotizing enterocolitis in 20,822 infants:
analysis of medical and surgical treatments. Clin Pediatr 2010;49:166–71.
take place with the clinical data available. It is critical to both adjust for
[14] Stey A, Barnert ES, Tseng CH, et al. Outcomes and costs of surgical treatments of
confounders and to include all the patients before and after the inter- necrotizing enterocolitis. Pediatrics 2015;135:e1190–7.
vention, not just evaluating the extra cost from the cases that result [15] Moss RL, Dimmitt RA, Barnhart DC, et al. Laparotomy versus peritoneal drainage
for necrotizing enterocolitis and perforation. N Engl J Med 2006;354:2225–34.
from NEC. This would more accurately help to account for the cost of
[16] Laparotomy vs. drainage for infants with necrotizing enterocolitis (NEST).
the intervention, as well as any other potential resource reduction (e.g. ClinicalTrials.gov Identifier NCT01029353.
decreased total parenteral nutrition days). More detailed modeling, [17] Ganapathy V, Hay JW, Kim JH, Lee ML, Rechtman DJ. Long term healthcare costs of
along with sensitivity analysis, can be done via decision analysis, taking infants who survived neonatal necrotizing enterocolitis: a retrospective longitudinal
study among infants enrolled in Texas Medicaid. BMC Pediatr 2013;13:127.
the opportunity to incorporate different at-risk populations, rates of [18] Berman L, Moss RL. Necrotizing enterocolitis: an update. Semin Fetal Neonatal Med
NEC, costs and incremental intervention benefits. These methods would 2011;16:145–50.
help separate out the cost of NEC removed from comorbidities that may [19] Dukhovny D, Pursley DM, Kirpalani HM, Horbar JH, Zupancic JA. Evidence,
quality, and waste: solving the value equation in neonatology. Pediatrics
or may not be related to the NEC diagnosis. Finally, the inclusion of 2016;137:e20150312.
costs beyond the neonatal hospitalization, parental out-of-pocket costs [20] Hodek JM, von der Schulenburg JM, Mittendorf T. Measuring economic con-
and other societal costs should be included to clearly understand the sequences of preterm birth – methodological recommendations for the evaluation of
personal burden on children and their caregivers. Health Econ Rev 2011;1:6.
total, long-term, impact of NEC. [21] Mowitz ME, Zupancic JA, Millar D, et al. Prospective economic evaluation along-
side the non-invasive ventilation trial. J Perinatol 2017;37:61–6.
[22] Spatz DL, Robinson AC, Froh EB. Cost and use of pasteurized donor human milk at a
418
children's hospital. J Obstet Gynecol Neonatal Nurs 2018;47:583–8. [26] Trang S, Zupancic JAF, Unger S, et al. Cost-effectiveness of supplemental donor
[23] Carroll K, Herrmann KR. The cost of using donor human milk in the NICU to achieve milk versus formula for very low birth weight infants. Pediatrics March
exclusively human milk feeding through 32 weeks postmenstrual age. Breastfeed 2018;141(3).
Med 2013;8:286–90. [27] O'Connor DL, Gibbins S, Kiss A, et al. Effect of supplemental donor human milk
[24] Ganapathy V, Hay JW, Kim JH. Costs of necrotizing enterocolitis and cost-effec- compared with preterm formula on neurodevelopment of very low-birth-weight
tiveness of exclusively human milk-based products in feeding extremely premature infants at 18 months: a randomized clinical trial. J Am Med Assoc
infants. Breastfeed Med 2012;7:29–37. 2016;316:1897–905.
[25] Arnold LD. The cost-effectiveness of using banked donor milk in the neonatal in- [28] Sanders GD, Neumann PJ, Basu A, et al. Recommendations for conduct, methodo-
tensive care unit: prevention of necrotizing enterocolitis. J Hum Lactat logical practices, and reporting of cost-effectiveness analyses: second panel on cost-
2002;18:172–7. effectiveness in health and medicine. JAMA 2016;316:1093–103.
419

Surgical considerations for neonates with necrotizing enterocolitis T

∗
Charles R. Hong, Sam M. Han, Tom Jaksic
Center for Advanced Intestinal Rehabilitation, Department of Surgery, Boston Children's Hospital and Harvard Medical School, Boston, MA, USA
Keywords: Necrotizing enterocolitis (NEC) is a potentially devastating condition that preferentially affects premature and
Necrotizing enterocolitis low birth weight infants, with approximately half requiring acute surgical intervention. Surgical consult should
Neonatal intensive care unit be considered early on, and deterioration despite maximal medical therapy or the finding of pneumoperitoneum
Prematurity are the strongest indications for emergent surgical intervention. There is no clear consensus on the optimal
Peritoneal drainage
surgical approach between peritoneal drainage and laparotomy; the best course of action likely depends on the
Laparotomy
Short bowel syndrome
infant's comorbidities, hemodynamic status, size, disease involvement, and available resources. Patients who
Neurodevelopment develop surgical NEC are at a significant risk for morbidity and mortality, with long-term complications in-
cluding short bowel syndrome, growth failure, and neurodevelopmental impairment. Further research into
strategies that optimize outcomes following surgery for NEC in the neonatal intensive care unit and long-term
are paramount.
1. Introduction and is generally associated with a lower risk for adverse outcomes
[4,5]. In addition, unlike NEC, virtually all patients who develop SIP
Necrotizing enterocolitis (NEC) is the most common surgical require surgical intervention. Studies that do not distinguish between
emergency in the neonatal intensive care unit (NICU), preferentially surgical NEC and SIP will be noted in the ensuing discussion.
affecting low birth weight and premature infants. The incidence of NEC
among very low birth weight (VLBW; birth weight 401–1500 g or ge- 2. Initial evaluation and management
stational age 22 weeks–29 weeks 6 days) neonates is 9% in North
America [1]. Approximately half of these infants require surgical in- Necrotizing enterocolitis exists as a spectrum of disease ranging
tervention, which is associated with a worse prognosis. Surgical NEC from an insidious onset of symptoms to rapid progression requiring
has an associated 35% mortality compared to 21% for medical NEC. surgical intervention. Signs and symptoms for NEC are nonspecific and
Moreover, the risk of NEC and mortality are inversely correlated with include abdominal distention, abdominal wall cellulitis, feeding intol-
birth weight (Fig. 1), with up to a 12% risk of NEC and 42% mortality in erance, emesis, bloody stools, temperature instability, lethargy, bra-
neonates with birth weight 500–750 g [2]. The pathogenesis of NEC dycardia, apnea, and hypotension. Biochemical indicators supporting a
remains unclear, but is thought to be multifactorial and related to im- diagnosis of NEC include acidosis, thrombocytopenia, and neutropenia
mature gut barrier, altered bacterial colonization, intestinal ischemia, [3].
and the consequent inflammatory response [3]. Care of the neonate Recent studies favor more refined diagnostic criteria for NEC as
with confirmed or suspected NEC requires interdisciplinary collabora- improved survival of premature newborns has led to a large cohort of
tion involving, but not limited to, neonatology, surgery, nursing, nu- confounding, non-NEC diagnoses [6]. Bell's criteria for diagnosing NEC
trition, pharmacy, and social work. This review will focus on the sur- has been used in evaluating severity of disease [7]. It has been modified
gical aspects of NEC management. since its inception in 1978 and includes three stages for diagnosing
It is important to note that in some studies of NEC, infants with NEC. Another perhaps less ambiguous classification system for diag-
spontaneous intestinal perforation (SIP) are grouped together with in- nosing NEC has been described and requires at least one clinical finding
fants who develop surgical NEC as it is not always possible to clinically and one radiographic finding to establish a diagnosis of NEC [8]. These
distinguish between the two conditions, particularly if primary perito- clinical findings include abdominal distention, gross blood in the stool,
neal drainage is pursued. Whereas SIP also affects premature infants, it and bilious gastric aspirate or emesis. Radiographic findings include
is a distinct clinical entity that tends to affect lower birth weight infants pneumatosis intestinalis, pneumoperitoneum, and portal venous gas.
∗
Corresponding author. 300 Longwood Avenue, Fegan 3, Boston, MA, 02115, USA.
E-mail address: tom.jaksic@childrens.harvard.edu (T. Jaksic).
C.R. Hong et al. Seminars in Fetal and Neonatal Medicine 23 (2018) 420–425
Fig. 1. Risk of necrotizing enterocolitis (NEC) and mortality from NEC by birth weight among very low birth weight infants (n = 71,808) in the Vermont Oxford
Network database 2005–2006 [2].
ultrasonography, had a significant association with both surgical in-

tervention and mortality [14]. These studies conclude that ultra-
sonography may be important in identifying high-risk infants who may
require surgical intervention, but the optimal timing and frequency of
ultrasonography are yet to be determined. Other limitations for ultra-
sonography use in NEC include operator variability, infant distress
while performing the study, poor visualization with intraluminal bowel
gas, and, at times, lack of 24 h availability.
In less severe cases, NEC is managed with bowel rest, gastric de-
compression, correction of hypotension and acidosis, along with par-
enteral nutrition. The prevention of sepsis with broad-spectrum anti-
biotics is imperative as sepsis has been associated with a significant
increase in morbidity and mortality within infants with NEC [3]. Serial
Fig. 2. Lateral radiograph demonstrating pneumoperitoneum in a neonate with abdominal radiographs, initially every 6–8 h, are taken to follow po-
necrotizing enterocolitis. tential disease advancement. Frequent clinical examinations are con-
ducted in infants with NEC, as bowel perforation, sepsis, and severe
decompensation may occur.
Plain abdominal radiographs play an important role in both the
diagnosis of NEC and in determining potential surgical intervention.
Pneumatosis intestinalis remains the hallmark radiographic finding in 3. Surgical indications for NEC
NEC; however, bowel wall thickening, fixed bowel loops, paucity of gas,
and portal venous gas are other imaging findings [9,10]. Pneumoper- Perforation or clinical deterioration after maximizing medical
itoneum, when seen on plain film radiography, requires emergent therapy are considered surgical indications for NEC. The presence of
surgical treatment (Fig. 2). The Duke Abdominal Assessment Scale pneumoperitoneum seen on radiography indicates perforation and is
(DAAS), a ten-point radiographic scale, was designed to assess the se- the only absolute criterion for surgery. However, studies demonstrate
verity of disease in infants with NEC [11]. The scale increases with that pneumoperitoneum may be present in less than half of NEC infants
disease severity, with a score of 0 given for normal loops of bowel and a with intestinal necrosis or perforation [10]. Portal venous gas and the
score of 10 for pneumoperitoneum. Patients with higher DAAS scores progression of pneumatosis intestinalis on serial abdominal radiographs
were more likely to undergo surgical intervention; however, ongoing indicate severe NEC, but neither are absolute indications for surgery.
criticism of its use exists due to the high observer variability in inter- Clinical findings that may suggest the potential need for surgery include
pretation of the different radiographic signs [11]. worsening abdominal girth over time, significant abdominal wall er-
Whereas plain film radiographs remain the reference standard ythema, and a fixed abdominal mass. Other important clinical factors
imaging modality for the diagnosis of NEC, ultrasonography has played include worsening bradycardia, apnea, temperature instability, and
an increasing role in evaluation of the disease. In the early stages of hypotension despite maximized medical therapy. One study of 35 in-
NEC, plain radiographs may only show nonspecific dilated loops of fants with confirmed NEC observed trends in seven different compo-
bowel; however, ultrasonography has been shown to identify bowel nents of metabolic derangement associated with NEC, including posi-
wall thickening, decreased perfusion of the bowel wall, portal venous tive blood cultures, acidosis, bandemia, neutropenia, hyponatremia,
gas, and pneumatosis intestinalis noted by the presence of hyperechoic hypotension, and thrombocytopenia [15]. Presence of three or more of
spots representing “air microbubbles” within the bowel wall that may the measured metrics was significantly associated with the need for
not be appreciable on radiography [9,12,13]. A recent meta-analysis surgical intervention. This predictive model is promising; however,
comprising 11 studies and 748 infants showed that focal fluid collec- prospective studies are needed to further characterize its validity.
tions, complex ascites, bowel wall thinning, absence of perfusion, bowel Accurately predicting the need for surgical intervention is important
wall thickening, and pneumoperitoneum, when seen on in improving overall outcomes. Present research has focused on
421
identifying new biomarkers that may differentiate infants with NEC preference, and available resources all influence the operative plan.
who will ultimately require surgical intervention. Fecal calprotectin, an Regardless of operative strategy, adequate resuscitation efforts to im-
inflammatory protein expressed by neutrophils and measured in stool, prove perfusion, correct acidosis, and reverse coagulopathy are integral
has been shown to be elevated in infants with NEC [16]. A calprotectin- to the care of these critically ill neonates.
related protein found in stool, S100A12, was also found to be sig- Peritoneal drainage for NEC was first described in the literature
nificantly elevated in infants with NEC who went on to develop bowel during the 1970s for five neonates with perforated NEC who were
perforation and receive surgery [6]. Although evidence suggests that considered too sick to undergo laparotomy with general anesthesia due
fecal calprotectin and S100A12 are elevated in infants with NEC, the to their small size, severe prematurity-associated comorbidities, and
specific cut-off values between studies were not similar, hence further overall critical illness [23]. The aim was to evacuate gas, pus, and feces
investigation is required. Moreover, the unpredictability of bowel via a small stab wound in the abdominal wall, with or without Penrose
movements and the availability of stool samples in this population drain placement. This would theoretically reduce peritoneal con-
limits the clinical applicability of these biomarkers. tamination and create a controlled fecal fistula, while allowing clinical
Intestinal fatty acid-binding protein (I-FABP) is found in small stabilization without the added stress of a more involved operation on
bowel enterocytes and is released into the blood following disruption. the infant. As the perforation may spontaneously seal over time, the
Elevated I-FABP has been linked to intestinal trauma and mesenteric fecal fistula would be allowed to close and enteral feedings attempted
ischemia in adult populations and has been shown to be a promising when clinically appropriate. Three of the five neonates survived beyond
biomarker for severe NEC [17]. A prospective cohort study showed their first year of life with normal intestinal function. Since then,
significantly higher I-FABP levels in both plasma and urine with a cut- peritoneal drainage has been increasingly used, either as primary
off value for diagnosing NEC at 9 ng/mL for I-FABP in plasma and therapy or as a temporizing measure to allow medical stabilization and
218 ng/mL in urine [18]. Results also showed a positive association delayed laparotomy. In a multicenter series in the USA that included
with higher plasma and urine I-FABP levels and complicated NEC re- SIP infants, 31% of VLBW infants with suspected surgical NEC or SIP
quiring surgery. Future validation studies of biomarkers for NEC underwent peritoneal drainage, with 46% of these infants also requiring
prognosis may allow for earlier identification of infants with NEC re- laparotomy [1]. Across all birth weight categories, neonates with sur-
quiring surgical intervention. gical NEC or SIP were more commonly treated with laparotomy;
It is important to consider late surgical indications for NEC related however, a higher percentage were treated with peritoneal drainage
to stricture formation. Post-NEC strictures are thought to occur as a among lower birth weight categories. The benefit of a delayed lapar-
consequence of ischemic intestinal injury and scarring following pre- otomy is it allows accurate assessment of the extent of diseased intes-
sentation. Retrospective series have reported a 17–24% rate of post- tine, though the downside is that it may be an unnecessary operation
NEC strictures among infants initially managed nonoperatively, com- for some infants. Overall, studies suggest that peritoneal drainage alone
pared to 24–36% among infants who required initial surgical inter- may be adequate for infants with SIP in most cases, while subsequent
vention [19,20]. Most of these occur in the colon or terminal ileum and laparotomy is often required in infants with NEC who undergo initial
can present with feeding intolerance, abdominal distention, or radio- peritoneal drainage [24,25].
graphic signs of obstruction. The development of strictures is difficult to Exploratory laparotomy is the most commonly utilized surgical
predict, with studies associating stricture formation with leukocytosis option, with the immediate goal of assessing extent of bowel injury,
and higher C-reactive protein levels during acute NEC, greater length of controlling peritoneal contamination, and resecting grossly necrotic
bowel resected, and failure to thrive. Thus, contrast studies should be bowel. The abdominal cavity is entered via a transverse supra-umbilical
considered when a neonate with a history of NEC develops obstructive incision, as this provides adequate exposure while allowing easier clo-
symptoms after initiation of enteral feeding. However, the rate of post- sure of the neonatal abdomen that is typically greater in width than
NEC strictures reported is higher when contrast studies are performed length. Frankly nonviable bowel is resected (Fig. 3), while bowel of
routinely, though not all are symptomatic; current practice favors close questionable viability is left in situ. It is important to preserve as much
clinical monitoring over routine contrast studies [21,22]. Unless stric- bowel length as possible to avoid resultant short bowel syndrome (SBS).
tures are symptomatic, they do not require surgical therapy. Placement of a silo with planned second-look procedure within 24–48 h
should be considered in situations where disease involvement is more
diffuse with multiple intervening segments of questionably viable
4. Surgical management of NEC bowel or when the patient is hemodynamically unstable [26]. Tradi-
tionally, the resection of perforated necrotic bowel for NEC is followed
Upon initial presentation of surgical NEC, there are multiple sur- by stoma creation with or without mucus fistula. The rationale is that,
gical options including primary peritoneal drainage, laparotomy with in the setting of perforation with gross contamination and likely he-
or without diverting ostomy, and silo placement. Patient hemody- modynamic instability, performing a primary anastomosis carries an
namics, comorbidities, weight, intraoperative findings, surgeon
Fig. 3. Exploratory laparotomy for necrotizing enterocolitis with frankly nonviable bowel and perforation.
422
increased risk for complications such as anastomotic leak or stricture. In clear consensus to guide optimal timing of ostomy reversal [39]. Earlier
selected patients with limited NEC, some authors have advocated for restoration of intestinal continuity may promote bowel adaptation and
primary anastomosis on the basis that no increased mortality or mor- reduce gastrointestinal losses, though the decision to pursue ostomy
bidity was observed, while stoma complications and additional surgery takedown should consider the infant's size, nutritional reserves, venti-
for re-establishment of intestinal continuity were avoided [27,28]. latory status, other comorbidities, and presence of stoma complications.
However, studies reporting on primary anastomosis during laparotomy Similar for other neonatal surgical conditions, families should also be
for NEC have been limited to retrospective reviews with small sample counseled prior to discharge on potential late complications following
sizes [29,30]. Bowel condition, as well as patient hemodynamics and surgical treatment of NEC, namely adhesive bowel obstructions, de-
comorbidities, should be considered when deciding upon stoma crea- layed anastomotic strictures, and anastomotic ulcers.
tion versus primary anastomosis. An alternative strategy of proximal
diverting ostomy without resecting bowel of indeterminate viability at 5. Long-term outcomes following surgical NEC
initial laparotomy for NEC has also been described [31]. This method
offers decompression and diversion of fecal stream from diseased per- Infants who survive following surgical intervention for NEC are at
forated bowel, while allowing spontaneous healing of diseased seg- risk for various long-term comorbidities. NEC with significant bowel
ments, liquefaction of necrotic segments, and presumably preservation loss can result in intestinal failure, more specifically SBS, in which re-
of more bowel length at time of ostomy takedown. However, studies sidual gastrointestinal absorptive function is inadequate to support
comparing this method to others are lacking. normal growth or maintain normal fluid and electrolyte homeostasis.
Studies comparing short term outcomes following initial peritoneal Aside from bowel resection, damaged bowel from NEC that is left in situ
drainage versus laparotomy for perforated NEC remain equivocal [32]. can also contribute to malabsorption or dysmotility. Multiple different
Two randomized multicenter studies, both including infants with SIP, criteria for intestinal failure are used in the literature; a commonly used
did not find a difference in survival or length of stay between initial definition is parenteral nutrition dependence for at least 90 days [40].
peritoneal drainage versus laparotomy [33,34]. Moss et al. randomized Necrotizing enterocolitis is the most common cause of neonatal
117 infants with birth weight < 1500 g to primary peritoneal drainage onset SBS, comprising up to 35% of diagnoses [41]. Infants who de-
versus laparotomy, with 38% of the drainage group receiving sub- velop SBS after NEC are at risk for various long-term complications,
sequent laparotomy; the authors concluded that the type of operation particularly those strongly associated with parenteral nutrition depen-
does not significantly influence 90-day mortality (35% for drainage vs dence such as intestinal failure associated liver disease, central line-
36% for laparotomy). Rees et al. randomized 68 infants with birth associated blood stream infections, and loss of vascular access. For-
weight < 1000 g and 74% of the drainage group underwent subsequent tunately, long-term survival of infants with SBS has significantly im-
laparotomy; no mortality differences were seen at one or six months proved over the past two decades with novel parenteral nutrition
and the authors concluded that primary peritoneal drainage was not an strategies and multidisciplinary intestinal rehabilitation programs
effective temporizing measure owing to high rates of delayed lapar- [42,43]. Children with SBS from NEC are also more likely to wean off
otomy. Of note, both randomized trials did not reach target enrollment parenteral nutrition relative to other causes [44], likely because infants
and thus were underpowered. Prospective cohort studies have sug- with NEC tend to be more premature and thus have a greater potential
gested that primary peritoneal drainage is associated with greater for growth and adaptation of their residual bowel. Nevertheless, chil-
mortality; however, there is often selection bias as infants undergoing dren with SBS due to NEC remain at risk for malabsorption and mi-
primary peritoneal drainage overall tend to be sicker with a higher cronutrient deficiencies after weaning off parenteral nutrition, as many
perceived operative risk [35]. In a multicenter prospective cohort study are missing their terminal ileum because of NEC.
in the USA, the mortality among VLBW infants with surgical NEC or SIP In the most severe cases of surgical NEC, there can be loss of the
who underwent peritoneal drainage alone was 50%, compared to entire small bowel, known as NEC totalis. This invariably leads to
31–34% among infants who underwent laparotomy alone or both “ultrashort” bowel syndrome with minimal chance for enteral au-
peritoneal drainage and laparotomy [1]. It is important to note that the tonomy. Whereas historically, an infant with NEC totalis has often been
group of patients who underwent peritoneal drainage alone likely re- directed towards comfort measures, modern home parenteral nutrition
presents a dichotomous group: one end of the spectrum consists of in- practices and multidisciplinary intestinal rehabilitation centers have
fants who undergo peritoneal drainage, then fail to improve and die made long-term PN safer and possible while maintaining acceptable
before they can receive a laparotomy, versus infants who improve fol- quality of life [45]. In addition, long-term PN can also serve as a bridge
lowing peritoneal drainage and never require laparotomy. The latter for eventual intestinal transplantation [46,47]. For the most extreme
subset likely consists of more infants with SIP or less severe surgical cases of surgical NEC, it is important to have a discussion with the fa-
NEC. mily that explores potential long-term options, while taking into con-
The early postoperative period after emergent surgical intervention sideration the infant's neurologic prognosis and other comorbidities.
for NEC is characterized by requirement of intensive hemodynamic and Aside from SBS, infants with a history of surgical NEC have greater
ventilatory support. Resuscitation efforts should pay close attention to healthcare needs and are at significant risk for impaired growth and
fluid and electrolyte shifts; neonates may have extensive losses of gas- neurodevelopment. This is likely due to a more severe disease process
trointestinal fluid with either nasogastric/orogastric decompression or in surgical NEC compounded upon prematurity and longer NICU stays.
high stoma outputs after resolution of ileus. Bowel rest and broad- From a multicenter North American database of extremely low birth
spectrum antibiotics should be continued, with reliance on parenteral weight (ELBW; birth weight 401–1000 g or 22–27 weeks gestational
nutrition. The optimal antibiotic regimen for NEC remains unclear and age) neonates, severe neurodevelopmental disability, as defined by bi-
warrants further research [36]. Hypergastrinemia with gastric acid lateral blindness, hearing impairment requiring amplification, inability
hypersecretion following extensive small bowel resection in neonates to walk 10 steps with support, cerebral palsy, and/or Bayley Mental or
has also been described; therefore, acid suppression with a proton pump Psychomotor Developmental Index < 70, was observed in 38% of
inhibitor or H2-blocker may be of benefit [37]. The provision of enteral children with a history of surgical NEC or SIP at 18–24 months cor-
nutrition is an interdisciplinary effort, and timing of initiating enteral rected age [48]. Similarly, the rate of severe growth failure, as defined
feeds postoperatively depends on the neonate's clinical stability, other by < 3rd percentile weight-for-age, among patients with a history of
comorbidities, stool/ostomy outputs, and intestinal anatomy. Common surgical NEC or SIP was 61% at discharge and 28% at 18–24 months
postoperative complications to be aware of include sepsis, intestinal corrected age [49]. Following discharge, infants with a history of sur-
stricture, wound infection, SBS, and those related to stomas [38]. gical NEC or SIP also more often undergo additional surgeries and have
Among infants who undergo stoma creation at laparotomy, there is no greater requirement for medical support such as oxygen use and tube
423
Fig. 4. Long-term outcomes of extremely low birth weight infants with necrotizing enterocolitis (NEC) at 18–24 months corrected age [48,49]. Surgical NEC includes
spontaneous intestinal perforation.
feeds (Fig. 4). It is important to guide family expectations regarding [4] Shah J, Singhal N, da Silva O, et al. Intestinal perforation in very preterm neonates:
risks for long-term morbidity within the surgical NEC population. risk factors and outcomes. J Perinatol 2015;35:595–600.
[5] Fisher JG, Jones BA, Gutierrez IM, et al. Mortality associated with laparotomy-
An ongoing question and active area of investigation pertains to the confirmed neonatal spontaneous intestinal perforation: a prospective 5-year mul-
effect of primary peritoneal drainage versus initial laparotomy on ticenter analysis. J Pediatr Surg 2014;49:1215–9.
neurodevelopmental outcomes. A prospective multicenter observa- [6] Gephart SM, Gordon PV, Penn AH, et al. Changing the paradigm of defining, de-
tecting, and diagnosing NEC: perspectives on Bell's stages and biomarkers for NEC.
tional cohort study of 156 ELBW neonates with surgical NEC or SIP Semin Pediatr Surg 2018;27:3–10.
compared the two treatments based on mortality or neurodevelop- [7] Bell MJ, Ternberg JL, Feigin RD, et al. Neonatal necrotizing enterocolitis.
mental impairment at 18–22 months [50]. Initial analysis favored la- Therapeutic decisions based upon clinical staging. Ann Surg 1978;187:1–7.
[8] Vermont Oxford Network. Manual of operations: Part 2. Data definitions and infant
parotomy, though there were no statistically significant differences data forms. Burlington, VT: VON; 2015. version 19.
after adjusting for confounders. The Necrotizing Enterocolitis Surgery [9] Coursey CA, Hollingsworth CL, Wriston C, Beam C, Rice H, Bisset 3rd G.
Trial (NEST) is an ongoing multicenter randomized trial that aims to Radiographic predictors of disease severity in neonates and infants with necrotizing
enterocolitis. Am J Roentgenol 2009;193:1408–13.
evaluate impairment-free survival at 18–22 months corrected age fol-
[10] Esposito F, Mamone R, Di Serafino M, et al. Diagnostic imaging features of necro-
lowing laparotomy versus peritoneal drainage for NEC or SIP. The high tizing enterocolitis: a narrative review. Quant Imag Med Surg 2017;7:336–44.
rates of long-term morbidity following surgical NEC mandates research [11] Markiet K, Szymanska-Dubowik A, Janczewska I, Domazalska-Popadiuk I,
efforts to assess for interventions that may improve long-term out- Zawadzka-Kepczynska A, Bianek-Bodzak A. Agreement and reproducibility of
radiological signs in NEC using the duke abdominal assessment scale (DAAS).
comes. Pediatr Surg Int 2017;33:335–40.
[12] Garbi-Goutel A, Brevaut-Malaty V, Panuel M, Michel F, Merrot T, Gire C. Prognostic
Practice points value of abdominal sonography in necrotizing enterocolitis of premature infants
born before 33 weeks gestational age. J Pediatr Surg 2014;49:508–13.
[13] McCarten KM. Ultrasound of the gastrointestinal tract in the neonate and young
• Up to half of infants who develop NEC require surgery. infant with particular attention to problems in the neonatal intensive care unit.
• Pneumoperitoneum and clinical deterioration despite maximal Ultrasound Clinics 2010;5:75–95.
[14] Cuna AC, Reddy N, Robinson AL, Chan SS. Bowel ultrasound for predicting surgical
medical therapy are the commonest indications for surgical inter- management of necrotizing enterocolitis: a systematic review and meta-analysis.
vention. Pediatr Radiol May 2018;48:658–66.
• Initial surgical options include peritoneal drainage and laparotomy; [15] Tepas 3rd JJ, Sharma R, Leaphart CL, Celso BG, Pieper P, Esquivia-Lee V. Timing of
surgical intervention in necrotizing enterocolitis can be determined by trajectory of
studies are ongoing to assess which may be the better option.
•
metabolic derangement. J Pediatr Surg 2010;45:310–3. discussion 313–14.
Surgical NEC is associated with greater mortality and morbidity [16] Aydemir O, Aydemir C, Sarikabadayi YU, et al. Fecal calprotectin levels are in-
than medical NEC. creased in infants with necrotizing enterocolitis. J Matern Fetal Neonatal Med
• Postoperative complications include intestinal stricture, SBS, and 2012;25:2237–41.

[17] Matsumoto S, Sekine K, Funaoka H, et al. Diagnostic performance of plasma bio-
neurodevelopmental impairment. markers in patients with acute intestinal ischaemia. Br J Surg 2014;101:232–8.
[18] Schurink M, Kooi EM, Hulzebos CV, et al. Intestinal fatty acid-binding protein as a
Research directions diagnostic marker for complicated and uncomplicated necrotizing enterocolitis: a
prospective cohort study. PLoS One 2015;10:e0121336.
[19] Heida FH, Loos MH, Stolwijk L, et al. Risk factors associated with postnecrotizing
• Long-term outcomes of initial peritoneal drainage versus lapar- enterocolitis strictures in infants. J Pediatr Surg 2016;51:1126–30.
[20] Phad N, Trivedi A, Todd D, Lakkundi A. Intestinal strictures post-necrotising en-
otomy.
• Timing of ostomy reversal after surgery for NEC.
terocolitis: clinical profile and risk factors. J Neonatal Surg 2014;3:44.
[21] Born M, Holgersen LO, Shahrivar F, Stanley-Brown E, Hilfer C. Routine contrast
• Optimal nutritional management following surgery for NEC. enemas for diagnosing and managing strictures following nonoperative treatment of
• Strategies to improve neurodevelopmental outcomes in patients necrotizing enterocolitis. J Pediatr Surg 1985;20:461–3.
[22] Gaudin A, Farnoux C, Bonnard A, et al. Necrotizing enterocolitis (NEC) and the risk
with surgical NEC. of intestinal stricture: the value of C-reactive protein. PLoS One 2013;8:e76858.
[23] Ein SH, Marshall DG, Girvan D. Peritoneal drainage under local anesthesia for
References perforations from necrotizing enterocolitis. J Pediatr Surg 1977;12:963–7.
[24] Cass DL, Brandt ML, Patel DL, Nuchtern JG, Minifee PK, Wesson DE. Peritoneal
drainage as definitive treatment for neonates with isolated intestinal perforation.
[1] Hull MA, Fisher JG, Gutierrez IM, et al. Mortality and management of surgical J Pediatr Surg 2000;35:1531–6.
necrotizing enterocolitis in very low birth weight neonates: a prospective cohort [25] Mishra P, Foley D, Purdie G, Pringle KC. Intestinal perforation in premature neo-
study. J Am Coll Surg 2014;218:1148–55. nates: the need for subsequent laparotomy after placement of peritoneal drains.
[2] Fitzgibbons SC, Ching Y, Yu D, et al. Mortality of necrotizing enterocolitis expressed J Paediatr Child Health Mar 2016;52:272–7.
by birth weight categories. J Pediatr Surg 2009;44:1072–5. discussion 1075–6. [26] Tan YW, Merchant J, Sharma V, et al. Extensive necrotising enterocolitis: objective
[3] Neu J, Walker WA. Necrotizing enterocolitis. N Engl J Med 2011;364:255–64. 20. evaluation of the role of second-look laparotomy in bowel salvage and survival.
424
World J Surg 2015;39:3016–22. [39] Struijs MC, Sloots CE, Hop WC, Tibboel D, Wijnen RM. The timing of ostomy closure
[27] Guelfand M, Santos M, Olivos M, Ovalle A. Primary anastomosis in necrotizing in infants with necrotizing enterocolitis: a systematic review. Pediatr Surg Int
enterocolitis: the first option to consider. Pediatr Surg Int 2012;28:673–6. 2012;28:667–72.
[28] Griffiths DM, Forbes DA, Pemberton PJ, Penn IA. Primary anastomosis for necro- [40] Duggan CP, Jaksic T. Pediatric intestinal failure. N Engl J Med 2017;377:666–75.
tising enterocolitis: a 12-year experience. J Pediatr Surg 1989;24:515–8. [41] Wales PW, de Silva N, Kim J, Lecce L, To T, Moore A. Neonatal short bowel syn-
[29] Downard CD, Renaud E, St Peter SD, et al. Treatment of necrotizing enterocolitis: an drome: population-based estimates of incidence and mortality rates. J Pediatr Surg
american pediatric surgical association outcomes and clinical trials committee 2004;39:690–5.
systematic review. J Pediatr Surg 2012;47:2111–22. [42] Merras-Salmio L, Pakarinen MP. Refined multidisciplinary protocol-based approach
[30] Haricharan RN, Gallimore JP, Nasr A. Primary anastomosis or ostomy in necrotizing to short bowel syndrome improves outcomes. J Pediatr Gastroenterol Nutr
enterocolitis? Pediatr Surg Int 2017;33:1139–45. 2015;61:24–9.
[31] Luzzatto C, Previtera C, Boscolo R, Katende M, Orzali A, Guglielmi M. Necrotizing [43] Puder M, Valim C, Meisel JA, et al. Parenteral fish oil improves outcomes in patients
enterocolitis: late surgical results after enterostomy without resection. Eur J Pediatr with parenteral nutrition-associated liver injury. Ann Surg 2009;250:395–402.
Surg 1996;6:92–4. [44] Sparks EA, Khan FA, Fisher JG, et al. Necrotizing enterocolitis is associated with
[32] Rao SC, Basani L, Simmer K, Samnakay N, Deshpande G. Peritoneal drainage versus earlier achievement of enteral autonomy in children with short bowel syndrome.
laparotomy as initial surgical treatment for perforated necrotizing enterocolitis or J Pediatr Surg 2016;51:92–5.
spontaneous intestinal perforation in preterm low birth weight infants. Cochrane [45] Mutanen A, Kosola S, Merras-Salmio L, Kolho KL, Pakarinen MP. Long-term health-
Database Syst Rev 2011(6):CD006182. related quality of life of patients with pediatric onset intestinal failure. J Pediatr
[33] Moss RL, Dimmitt RA, Barnhart DC, et al. Laparotomy versus peritoneal drainage Surg 2015;50:1854–8.
for necrotizing enterocolitis and perforation. N Engl J Med 2006;354:2225–34. 25. [46] Sanchez SE, Javid PJ, Healey PJ, Reyes J, Horslen SP. Ultrashort bowel syndrome in
[34] Rees CM, Eaton S, Kiely EM, Wade AM, McHugh K, Pierro A. Peritoneal drainage or children. J Pediatr Gastroenterol Nutr 2013;56:36–9.
laparotomy for neonatal bowel perforation? A randomized controlled trial. Ann [47] Vennarecci G, Kato T, Misiakos EP, et al. Intestinal transplantation for short gut
Surg 2008;248:44–51. syndrome attributable to necrotizing enterocolitis. Pediatrics 2000;105:E25.
[35] Sola JE, Tepas 3rd JJ, Koniaris LG. Peritoneal drainage versus laparotomy for ne- [48] Fullerton BS, Hong CR, Velazco CS, et al. Severe neurodevelopmental disability and
crotizing enterocolitis and intestinal perforation: a meta-analysis. J Surg Res healthcare needs among survivors of medical and surgical necrotizing enterocolitis:
2010;161:95–100. a prospective cohort study. J Pediatr Surg 2017;53:101–7.
[36] Shah D, Sinn JK. Antibiotic regimens for the empirical treatment of newborn infants [49] Hong CR, Fullerton BS, Mercier CE, et al. Growth morbidity in extremely low birth
with necrotising enterocolitis. Cochrane Database Syst Rev 2012;8:CD007448. weight survivors of necrotizing enterocolitis at discharge and two-year follow-up.
[37] Hyman PE, Everett SL, Harada T. Gastric acid hypersecretion in short bowel syn- J Pediatr Surg 2018;53:1197–202.
drome in infants: association with extent of resection and enteral feeding. J Pediatr [50] Blakely ML, Tyson JE, Lally KP, et al. Laparotomy versus peritoneal drainage for
Gastroenterol Nutr 1986;5:191–7. necrotizing enterocolitis or isolated intestinal perforation in extremely low birth
[38] Horwitz JR, Lally KP, Cheu HW, David Vazquez W, Grosfeld JL, Ziegler MM. weight infants: outcomes through 18 months adjusted age. Pediatrics
Complications after surgical intervention for necrotizing enterocolitis: a multicenter 2006;117:e680–7.
review. J Pediatr Surg 1995;30:994–9.
425

Neurodevelopmental outcomes following necrotizing enterocolitis T

a b a,∗
Marie Hickey , Michael Georgieff , Sara Ramel
a
Department of Pediatrics, Division of Neonatology, University of Minnesota, Minneapolis, MN, USA
b
Department of Pediatrics and Center for Neurobehavioral Development, Division of Neonatology, University of Minnesota, Minneapolis, MN, USA
Keywords: Necrotizing enterocolitis (NEC), a gastrointestinal emergency predominantly affecting premature infants, is
Necrotizing enterocolitis associated with increased risk for poor neurodevelopmental outcomes. NEC often strikes during a period of rapid
Outcomes and dynamic neurologic development when the brain is particularly vulnerable to insults and nutrient deficits.
Neurodevelopment The pathogenesis of neurodevelopmental impairment following NEC is likely multifactorial, with both nutri-
Prematurity
tional and non-nutritional factors at play. Follow-up testing that ensures early detection and intervention for
Nutrition
Inflammation
impairments is crucial to optimize neurodevelopmental outcomes following NEC. A multifaceted approach to
Gastrointestinal microbiota follow-up after NEC is necessary, with close monitoring of growth, serial developmental assessments, neurologic
examinations, hearing and vision testing and neuroimaging. Further research is needed to understand the pa-
thogenesis of neurodevelopmental impairment following NEC, to identify more targeted follow-up tests, and to
discover interventions aimed at optimizing neurodevelopmental outcomes following NEC.
1. Introduction delays are therefore of utmost importance to optimize outcomes for this
high-risk population. This article begins by reviewing the evidence that
Necrotizing enterocolitis (NEC) is a common acquired life-threa- NEC is associated with increased risk for neurodevelopmental impair-
tening gastrointestinal condition in newborns. It primarily affects pre- ment. Then we discuss the mechanisms that likely mediate this re-
term born infants, occurring in approximately 10% of extremely low lationship. Finally, we explore methods of neurodevelopmental mon-
birth weight infants (ELBW). Surgery is required in 50–70% of these itoring after NEC.
infants [1]. Despite advances in antenatal and perinatal care and in-
creased survival rates for very low birth weight (VLBW) infants, NEC 2. NEC and neurodevelopmental impairment: a review of the
remains a potentially devastating complication for this population of recent literature
infants. NEC is not only an important intestinal emergency, but it also
has broader sequelae including systemic inflammation, hypoxia, The relationship between NEC and neurodevelopmental impairment
ischemia, and multisystem organ failure. It can also have lasting effects has been documented in numerous studies and systematic reviews.
beyond the acute period, including short bowel syndrome, malnutri- Several recent small matched control studies demonstrate a significant
tion, re-hospitalizations, and neurodevelopmental delay. association between NEC and neurodevelopmental impairment.
Perinatal brain injury remains a common cause of significant long- Soraisham et al. found that infants who develop stage II or III NEC were
term impairment in preterm infants. This injury can occur through a at significantly higher risk for the development of neurodevelopmental
variety of mechanisms including, but not limited to, inflammation, disability when compared to age-matched controls at 36 weeks cor-
hypoxic/ischemic injury, hemorrhage secondary to coagulopathy, and rected gestational age [2]. Dilli et al. demonstrated significantly lower
toxin exposure. Unfortunately, infants who develop NEC often ad- median mental developmental index (MDI) and psychomotor develop-
ditionally experience many of these complications and suffer the con- mental index (PDI) scores in a small cohort of VLBW infants with NEC
sequences of long-term neurodevelopmental impairment as a result. compared to controls [3]. In a small cohort of VLBW survivors of NEC,
Attempts at preventing NEC and its associated brain injury have not Sonntag et al. found significant neurodevelopmental delay at 12 and 20
been particularly effective. Improved understanding of the mechanisms months compared to age-matched controls without NEC [4]. Salhab's
behind this association, along with early detection and intervention for group reported that a small cohort of ELBW infants with NEC had
∗
Corresponding author. Department of Pediatrics, Division of Neonatology University of Minnesota, East Building, MB630, 2450 Riverside Ave, Minneapolis, MN,
55454, USA.
E-mail address: Sramel@umn.edu (S. Ramel).
M. Hickey et al. Seminars in Fetal and Neonatal Medicine 23 (2018) 426–432
significantly lower mean PDI at 18 months corrected age compared to outcome.

matched controls. There were no differences in MDI between groups, Whereas the current literature demonstrates a convincing associa-
nor were there MDI or PDI differences between stage II and stage III tion between NEC and increased risk for poor neurodevelopmental
NEC in this study [5]. outcomes, the specific causes underlying neurodevelopmental impair-
These findings of neurodevelopmental impairment following NEC ment following NEC remain to be clarified. Further understanding of
from small case–control studies are confirmed in larger, multi-center the mechanisms behind these associations is an important first step in
analyses, including a systematic review of observational studies re- developing interventions to improve long-term outcomes for this vul-
porting on neurodevelopmental outcomes of VLBW survivors after NEC nerable population.
performed by Schulzke et al. This review reported that the risk of
neurodevelopmental impairment after one year corrected age was sig- 3. Proposed mechanisms underlying associations between NEC
nificantly higher in infants with NEC compared to infants who were not and neurodevelopment
diagnosed with NEC [6]. Another systematic review by Rees et al. found
that, when compared to age- and gestation-matched infants without Late gestation and the first several years of life are periods of rapid
NEC, infants with NEC were significantly more likely to be neurode- and ongoing brain development that have been identified as “sensitive
velopmentally impaired, specifically having increased risk of cerebral periods” of time to optimize development. The brain is composed of
palsy, visual, cognitive and psychomotor impairment [1]. heterogeneous anatomical regions and functions, each with their own
There is more inconsistency among studies when examining the unique “critical period” for development. Each individual region must
question of whether infants with surgical versus medical NEC have develop appropriately, in a time-sensitive manner, to achieve co-
poorer neurodevelopmental outcomes. Dilli's group did not find sig- ordinated development of brain areas that work synergistically to carry
nificant differences in neurodevelopmental outcomes between NEC out complex functions [11]. The second and third trimesters are marked
survivors who required surgery and those with medically treated NEC by rapid brain development including neurogenesis, neuronal migra-
[3]. Similarly, in a large, single-center retrospective analysis, Shah's tion, maturation, apoptosis and synaptogenesis [12]. Infants born pre-
group found that ELBW infants who develop NEC and spontaneous term may therefore be particularly vulnerable to factors that disrupt
intestinal perforation (SIP) are at increased risk for neurodevelop- these processes, including malnutrition, impaired perfusion and oxygen
mental impairment measured by Bayley testing, and there were no supply, and cytotoxic mediators that accompany inflammatory pro-
significant differences in neurodevelopmental outcomes observed be- cesses. As a disease process that primarily affects preterm infants, the
tween medical and surgical NEC groups in that study [7]. In contrast, evidence of a significant relationship between NEC and neurodevelop-
Hintz et al. showed increased risk for adverse neurodevelopmental mental disability is not surprising. It is likely that this association is
outcomes among infants with surgical NEC but not in infants with multifactorial, with both nutritional and non-nutritional insults ac-
medically managed NEC in a large retrospective analysis. Infants with companying the intestinal manifestations of NEC and causing harm to
surgical NEC demonstrated increased risk for MDI < 70 and PDI < 70 the developing brain (Fig. 1). Furthermore, the timing, severity and
on Bayley testing and the combined outcome of neurodevelopmental duration of these insults will determine the specific locations of brain
impairment at 18–22 months corrected age compared with infants who injury and the functional outcomes.
were not diagnosed with NEC [8]. The previously mentioned studies by
Rees and Schulzke also found that surgical NEC is associated with 4. Nutritional factors affecting neurodevelopment after NEC
greater risk for neurodevelopmental impairment compared to medical
NEC [1,6]. Nutrition plays a critical role in brain development and remains one
Culture-confirmed infection in the neonatal period is associated of the few factors that the care team can manipulate. Nutrient intake
with increased rates of adverse neurodevelopmental outcomes, and recommendations for stable preterm infants are well established;
infants with NEC have a significantly higher incidence of culture-con- however, the optimal nutritional management of infants with NEC is
firmed sepsis [9]. Martin et al. explored this relationship in a large not well described. Infants who are diagnosed with NEC are at sig-
prospective cohort study of infants included in the Extremely Low Ge- nificant risk for undernutrition for multiple reasons. Not only do they
stational Age Newborns (ELGAN) Study. They reported increased risk of often have compromised nutritional intake, but they may also have
neurodevelopmental impairment, cerebral palsy and microcephaly in persistent mucosal injury following NEC that can lead to decreased
children who had surgical NEC, especially if they also had late bac- absorption of nutrients. These infants often have a prolonged period
teremia [10]. Several additional studies support this association be- without enteral feedings during the acute phase of the illness.
tween NEC and sepsis [2,5,7,8]. Other common complications of pre- Parenteral nutrition is the primary form of nutrition while enteral
maturity are reported to be associated with NEC, including longer feedings are held. NEC and its treatment have the potential to induce
intubation [5], chronic lung disease [2,8], intraventricular hemorrhage fundamental changes in energy, protein and micronutrient metabolism
[7], retinopathy of prematurity [7], and longer hospital stays [2,5]. that upset nutrient balance. Hyperglycemia is common and necessitates
Whether a result of NEC, or an indication of the individual infant's level a restricted glucose infusion rate. Lipid intake, a significant source of
of vulnerability or illness above and beyond NEC, these additional calories when on intravenous nutrition, is often also often limited due
complications compound risk for neurodevelopmental impairment. to cholestasis related to intestinal failure, ascending cholangitis, and
Many of the studies investigating neurodevelopmental outcomes bile sludging following prolonged parenteral nutrition.
following NEC are limited by small numbers of patients from single Whereas there is a paucity of research in preterm infants, adults
centers. Furthermore, meta-analyses carry the challenge of generalizing who are septic demonstrate a requirement for higher energy and pro-
the results of small individual observational studies with varied study tein delivery due to increased cellular oxygen consumption and nega-
designs, patient populations, severity of illness, management, method tive nitrogen balance [13]. Protein status is of particular importance
of follow-up and age at follow-up. Despite these limitations, the re- due to its role in fat-free mass (FFM) accretion, neurogenesis, and
cently published literature makes a persuasive argument for the asso- neuronal differentiation [14]. Protein has been shown to play a role in
ciation between NEC and worsened neurodevelopmental outcomes stimulating neuronal growth factors including insulin-like growth
among preterm infants (Table 1). Efforts to stratify risk for neurode- factor-1 (IGF-1). Increased levels of IGF-1 have been associated with
velopmental impairment by specific factors associated with NEC, in- lower risk for abnormal cognition in preterm infants [15]. In one multi-
cluding severity of illness, complications associated with NEC, and need center cohort study of infants with surgical NEC, Lin et al. reported that
for surgical intervention lack consensus. Large prospective studies with high protein provision in infants with surgical NEC was associated with
long-term follow-up are clearly needed to gauge the risk to long-term increased head circumference, an index of brain growth [16]. Recent
427
M. Hickey et al.
Table 1
Neurodevelopmental outcomes among preterm infants.
Study Study design Population sample/size NDI measurement Age at follow-up Conclusions
scale
Sonntag et al. [4] Matched control 20 of 22 surviving VLBW infants Griffiths 12 and 20 months corrected Neurodevelopment significantly delayed in infants with NEC. 55% in NEC group
age had severe developmental delay compared to 22.5% of infants without NEC
Salhab et al. [5] Matched control 17 ELBW infants with NEC BSID 2 18 and 22 months corrected Compared to controls, infants with NEC had lower PDI and higher incidence of
age abnormal neurologic examination
Soraisham et al. Matched control 51 VLBW infants with NEC BSID 2 and 36 months corrected age 24% of infants with NEC had one major neurodevelopmental disability compared
[2] Sanford–Binet with 10% among control infants.
NEC group had significantly higher incidence cognitive index < 70 and visual
impairment
Dilli et al. [3] Matched control 20 of 39 surviving VLBW infants with NEC BSID 2 18 and 24 months corrected Lower MDI and PDI in infants with NEC compared to controls;
age Neurodevelopmental outcomes did not differ between NEC survivors with and
without surgery
428
Hintz et al. [8] Multicenter, retrospective 2948 ELBW infants from NICHD Research BSID 2 18 and 20 months corrected Surgical but not medical NEC was associated with adverse neurodevelopmental
analysis Network (245 of whom had NEC) age outcomes, including MDI < 70 and PDI < 70, at 18 and 22 months corrected age
compared to no NEC
Schulzke et al. [6] Systematic review VLBW infants in 11 randomized studies including Bayley or Griffiths ≥1 year Risk of NDI significantly higher in infants with NEC compared to no NEC; risk
five studies with matched controls scale even higher in surgical vs medically managed NEC
Rees et al. [11] Systematic review 7843 VLBW and ELBW infants (821 with NEC) BSID 2 and Griffiths Median follow-up at 20 Infants with NEC were significantly more likely than matched controls without
months corrected age NEC to have NDI including cerebral palsy, visual, cognitive and psychomotor
impairment; risk for neurological impairment significantly increased in stage III
and surgical NEC
Shah et al. [7] Population-based 1722 ELBW infants (208 with NEC) in Cincinnati BSID 2 Hospital discharge, 18 and Increased risk of NDI compared to controls; no differences in neurodevelopmental
retrospective analysis Collaborative Outreach Program Database of 22 months corrected age outcomes between medical NEC, surgical NEC and SIP groups
NICHD NRN
NDI, neurodevelopmental impairment; VLBW, very low birth weight; NEC, necrotizing enterocolitis; ELBW, extremely low birth weight; BSID, Bayley Scales of Infant Development; PDI, psychomotor developmental
index; MDI, mental developmental index; NICHD NRN, National Institute of Child Health and Human Development Neonatal Research Network; SIP, spontaneous intestinal perforation.
Fig. 1. Proposed multifactorial pathogenesis underlying neurodevelopmental impairment after necrotizing enterocolitis. BM, breast milk; OFC, oral feed challenge;
FFM, fat-free mass.
research findings support the importance of protein in brain growth, infants. Studies using body composition measurements reveal that the
but the value of non-protein macronutrients and micronutrients should quality of growth in preterm infants is altered, with decreased amounts
not be discounted. Iron and zinc are two of the many micronutrients of FFM and increased adiposity noted at term corrected age, specifically
that are at risk following NEC. The absorption and trafficking of iron are in those infants who were more critically ill [14,21,23,24]. Because of
significantly altered in the presence of inflammation because of the the pervasive nature of postnatal growth failure in preterm infants, its
activation of hepcidin, and zinc sufficiency is jeopardized following consequences, including associations with poorer neurodevelopmental
NEC due to decreased intestinal absorption. The consequences of iron outcomes, have been well documented [22,25,26].
and zinc deficiency are abnormal neurodevelopment and poor growth Compared to their healthy counterparts, sick neonates are at even
[13]. These macro- and micronutrient deficits following NEC interfere higher risk for growth failure and neurodevelopmental impairment
with somatic growth and brain development during a critical period in [13]. The underlying etiology of growth failure and risk for poor neu-
this population of infants. rodevelopmental outcome following NEC is likely multifactorial. Diffi-
Given the importance of nutrients in brain development and the culty feeding during the acute stage of the illness and resultant mal-
high risk for nutrient deficiencies and growth failure following NEC, nutrition seems to provide a credible but incomplete explanation. Both
further studies with the aim of optimizing nutritional management after inadequate nutritional delivery and other non-nutritional mechanisms
NEC are necessary. These might include evaluation of increased protein such as inflammation and key changes in cellular metabolism that alter
provision following NEC. Existing studies of nutritional strategies that nutrient demand, availability, and handing are likely also at play in the
optimize administration of high-quality protein in preterm infants have observed changes in growth and development [13,27]. A series of
demonstrated safety and improved outcomes [17,18]. Characterization physiological changes, involving inflammatory factors and hormones,
of the ideal timing and duration of adjustments in nutrient supply to in response to the stress of an acute illness promotes short-term survival
achieve continued growth and optimize neurodevelopment following but is likely detrimental to long-term growth and development [21].
NEC would also be helpful. Alterations in the growth hormone (GH) axis have been suggested in the
relationship between illness and risk for poor growth and neurodeve-
5. Non-nutritive factors affecting neurodevelopment after NEC lopmental outcomes. Critically ill children become relatively GH re-
sistant and demonstrate elevated GH levels with inappropriately low
5.1. Growth and neurodevelopment IGF-1 levels [20,28,29]. As discussed previously, higher IGF-1 levels
have been associated with improved neurodevelopmental outcomes
Despite attempts to optimize nutritional provision, growth failure, [15].
even in preterm infants without NEC, has been well demonstrated. As Although we may be able to extrapolate some findings from studies
many as 79% of VLBW premature infants remain below the 10th per- of other illnesses, follow-up studies on long-term growth and its asso-
centile in weight at 36 weeks post-conceptual age [19]. As nutrition, ciation with neurodevelopmental outcomes specifically following NEC
growth and relationships to neurodevelopmental outcomes have been are few, and the results are inconsistent. Hintz et al. reported significant
more extensively studied, the utility of growth parameters in addition growth failure only in ELBW infants with NEC requiring surgical
to weight have been assessed. Linear growth has been identified as an treatment but not in infants treated medically for NEC compared to
important indicator of FFM accretion and a potential marker of future controls. Two other studies did not find any significant differences in
cognitive deficit in VLBW preterm infants [20]. Even in infants without weight, length, or head growth between NEC survivors and control
NEC, linear growth failure persisting beyond the second year of life is infants without NEC [3,4]. Further studies of growth, inflammation,
common [21,22]. Accurate linear growth measurements are difficult to and hormonal changes are warranted to better understand the role of
obtain, and, more recently, some centers have used air-displacement non-nutritional factors in growth and neurodevelopmental outcomes
plethysmography to non-invasively measure body composition in following NEC.
429
5.2. Inflammation and neurodevelopment underlying this relationship, including changes in expression of genes
and proteins involved in neurotransmission, synaptic plasticity and
Increased levels of local tissue and circulating pro-inflammatory metabolism, control of stress hormones, or neuronal signaling via the
cytokines are involved in the pathogenesis of NEC. In one hypothesis, vagus nerve [42]. Further studies are needed to determine whether
intestinal injury and barrier dysfunction permit translocation of in- these relationships exist in humans and to explore the functional con-
flammatory mediators and bacteria into systemic circulation, resulting sequences of these potentially important physiological changes asso-
in a systemic inflammatory response. The observation that increased ciated with an altered gut microbiota.
disease severity (i.e. more significant injury to the intestinal barrier and
resultant systemic inflammation) is associated with adverse neurode- 5.4. Other factors
velopment supports this theory [7]. Ubiquitous signaling molecules
with pro-inflammatory, anti-inflammatory, trafficking, growth, and The observation that infants with surgical NEC have increased risk
damage repair roles, including cytokines, chemokines, and growth of poor neurodevelopmental outcomes raises the question of whether
factors are implicated in the relationship between infections distant this association could be related to a heightened inflammatory response
from the brain and brain injury [30]. In a multicenter study of infants following surgery, the possibility of anesthetic-related neurotoxicity, or
affected by NEC, increased serum levels of pro-inflammatory cytokines the severity of the disease process (need for surgery indicating ad-
were found to be associated with increased risk for poor growth and vanced NEC). In addition, the type of surgery performed may play a role
development [31]. A similar phenomenon occurs with maternal and/or in long-term outcomes, including neurodevelopment. In a multicenter
neonatal inflammation or infection, as these infants have also been cohort study comparing initial laparotomy and peritoneal drain place-
shown to be at higher risk for long-term neurodevelopmental impairment in ELBW infants with NEC, a risk-adjusted odds ratio for death or
ments, in particular cerebral palsy, in several cohort studies and a meta- neurodevelopmental impairment favored laparotomy over peritoneal
analysis of these studies [32]. drain placement [43]. This may be due to the potential for needing a
White matter injury, often manifested as periventricular leukoma- second surgery, prolonged courses of nil-per-os and/or antibiotics or
lacia (PVL), is an important example of disrupted brain development other undetermined factors. Additional studies comparing laparotomy
and neurologic disability in preterm infants [33]. The underlying pa- and peritoneal drain placement focus on outcomes of mortality, rates of
thogenesis of cerebral white matter injury involves cerebral ischemia, dependence on enteral feeding and duration of hospital say and are
systemic infection or inflammation, and intrinsic vulnerability of cer- inconclusive. Large, multicenter randomized controlled trials including
ebral white matter pre-myelinating oligodendrocytes, particularly neurodevelopmental and growth outcomes are needed. The question of
around 28–32 weeks gestation. Also peaking during this high-risk ge- anesthetic-related neurotoxicity is also unsettled. In animal models,
stational period is cerebral microglial activation, which becomes anesthetics have been found to increase neuronal cell death following
heightened in the context of systemic inflammation and contributes to exposure. Subsequent neurologic abnormalities have been noted in
ongoing injury [34]. Observational studies have shown an association some studies. Translation of these findings from animals to humans is
between PVL and maternal intrauterine infection with fetal systemic problematic and observational studies of children exposed to surgery
inflammation or neonatal systemic infection with neonatal systemic are conflicting [44].
inflammation [32,35,36]. Adding yet another facet to the hypothesis of a multifactorial pro-
cess mediating the relationship between NEC and neurodevelopment, it
5.3. Intestinal microbiota, inflammation, and brain development has been proposed that NEC pathophysiology may only be a small part
of, or simply a maker for, other risk factors. For example, the bowel
The role of the intestinal microbiota in the prevention, pathogen- immaturity that predisposes an infant to NEC may be an indicator of
esis, and potentially the neurodevelopmental consequences following brain maturity and vulnerability beyond that attributed to gestational
NEC deserves mentioning here. The gut microbiota, a diverse and age. NEC may simply be a marker of illness severity during an infant's
complex community composed of trillions of microbes, is known to play NICU stay [7]. Regardless of the exact role that NEC itself plays in the
an important role in the developmental programming of gut epithelial risk for poor neurodevelopmental outcomes, infants who are affected by
barrier function and the gut immune system. Recently, new studies this disease require close follow-up.
have shown an important link between the microbes that inhabit the
intestinal tract and the developing brain [37]. 6. Neurodevelopmental follow-up after NEC
Knowledge of the factors that influence the neonatal gut microbiota
has greatly expanded in the last few decades. The intestinal microbiota While studies in the prevention of NEC and its long-term sequelae
of the neonate appears to be highly sensitive to maternal health status, are ongoing, the interventions with the highest likelihood of improving
antenatal infection, antibiotic exposure and preterm birth [37]. The outcomes related to this potentially devastating disease remain in vig-
infant's mode of delivery, type of nutrition and early use of antibiotics ilant follow-up, monitoring and serial assessments with the goal of early
alter the composition of the gut microbiota and may have long-lasting detection of impairments and prompt intervention.
effects [38]. Exposure to antibiotics early in life has been shown to
significantly alter the development of the neonatal gut microbiota. In 6.1. Clinical monitoring and neuropsychological assessment
fact, there is evidence that antibiotic treatment causes more drastic and
sustained changes to the gut microbiota when compared to other clin- Although they are not a direct measure of brain development, an-
ical factors [39–41]. thropometric measurements, including weight gain, linear growth and
The local damage to the intestine, withholding of enteral feedings head circumference, are important components of follow-up for a pre-
and antibiotic treatment that accompany NEC all pose significant risk term infant affected by NEC. Linear growth reflects lean body mass and
for a perturbed gut microbiota in the infants who suffer from this dis- protein accretion. It correlates with brain growth, making it an im-
ease process. An altered intestinal microbiota may intensify in- portant biomarker to predict long-term developmental outcomes in
flammation, potentiate intestinal barrier dysfunction and delay healing. VLBW infants [21]. Studies in the use of body composition in preterm
Furthermore, a perturbed gut microbiota has the potential to influence infants show that increased early FFM gains are associated with im-
neurodevelopmental outcomes via inflammation and possibly has a proved neurodevelopmental outcomes, including higher cognitive and
direct effect on the central nervous system. Recent animal model stu- motor scores, faster speed of processing, improved working memory,
dies exploring the link between an altered gut microbiota and brain and higher IQ [24]. Since the “critical period” of brain development
development have uncovered evidence for potential mechanisms extends until about three years of age, correction of growth failure in
430
the first few months to years of life, or better yet prevention, can im- through physical exam, developmental testing, and other imaging, re-
prove neurodevelopmental outcomes. Frequent monitoring of easily mains to be elucidated. Before this becomes standard practice, there
obtainable growth measures can ensure identification and prompt in- needs to be more clear determination of practical considerations such as
terventions with a higher likelihood of optimizing neurodevelopmental ability to acquire images without sedating medications. In addition,
outcomes. whereas abnormal findings are often associated with long-term dis-
The Bayley Scales of Infant Development (BSID) is a widely used ability, some infants develop without significant disability despite ab-
tool in assessment for neurodevelopmental delay. This standardized normalities on neuroimaging. Neuroimaging may serve as a helpful tool
developmental test is important in early detection of developmental in predicting neurodevelopmental outcome, but long-term follow-up to
delay; however, in determining need for early intervention programs understand the evolution of brain injury over time and assess functional
and evaluating treatment effectiveness, it is not without limitations. outcomes is necessary.
Poor predictive value of early testing for cognitive impairment at school
age and underestimation of delays are cited limitations of the BSID-II 7. Conclusions
and BSID-III, respectively [45,46]. These assessments are best applied
serially and used in conjunction with regular thorough neurologic ex- In summary, NEC remains an important, life-threatening illness that
aminations to assess the quality of motor skills, coordination, gait and primarily affects preterm infants and can have life-long complications.
behavior in addition to hearing and vision evaluation by appropriate Infants with NEC are at increased risk for poor neurodevelopmental
professionals. outcomes, a relationship that is likely multifactorial, including nutri-
One aspect of brain development that is likely to be affected by tional and non-nutritional factors. Survivors of NEC require long-term
prematurity, poor nutrition and growth is speed of neuronal processing, follow-up to monitor for signs of neurodevelopmental impairment to
due to effects of these insults on the processes of synaptogenesis and ensure prompt intervention.
myelination. Speed of processing can be assessed as soon as early in-
fancy with visual evoked potentials (VEPs), which are time-locked 7.1. Practice points
electroencephalographic recordings representing the brain's response to
a specific visual stimulus. Although VEP is not currently used broadly in • Current evidence indicates that the diagnosis of NEC is associated
non-research settings, it may give some insight into early postnatal with increased risk for neurodevelopmental impairment; for infants
brain development and provide a clearer connection between neonatal with a history of NEC, long-term follow-up, early diagnosis and in-
risk factors and brain function. More research is needed to identify tervention are critical to improving and optimizing long-term neu-
specific testing modalities that can reliably identify impairments in the rodevelopmental outcomes.
areas of the brain that are still developing, and therefore particularly • There is not one follow-up test that alone can reliably predict risk for
vulnerable at the time when NEC and its associated insults occur. neurodevelopmental impairment currently; instead, a combination
of testing methods provides the best risk stratification.
6.2. Neuroimaging • Most neurodevelopmental follow-up studies are performed at young
ages and may not represent true long-term outcomes; longitudinal
Cranial US (CUS) is currently the most widely used neuroimaging follow-up beyond preschool age is necessary.
technique in this population. It is an ideal mode of imaging to detect
intraventricular hemorrhage and cystic PVL; however, it has limited 7.2. Research directions
value in detecting diffuse white matter injury in studies comparing
neonatal ultrasonography with magnetic resonance imaging (MRI) • Randomized trials aimed at improving nutrition, decreasing sys-
[47,48]. The role of MRI in neonatal care has been increasingly in- temic inflammation and preserving the intestinal microbiota would
vestigated in the last decade as a potential tool to assist in early prog- promote improved understanding of the pathogenesis underlying
nostic evaluation. the relationship between NEC and poor neurodevelopmental out-
Several studies evaluating the prognostic power of MRI for predic- comes, and inform possible preventative measures.
tion of neurodevelopmental abnormalities have been performed in • There is a need for prospective, longitudinal follow-up studies after
preterm infants, a population with high risk for neurodevelopmental NEC, exploring specific assessments of brain areas that are known to
impairment (even in the absence of NEC). Both late CUS and near-term be developing rapidly in the third trimester to assist in the precise
MRI abnormalities were associated with adverse neurodevelopmental targeting of interventions to optimize outcomes.
outcomes at 18–22 months corrected age in a cohort of infants with
birth gestational age < 28 weeks in the Neonatal Research Network. Conflicts of interest
Early CUS findings were not associated with adverse outcomes. This
same study also reported that a considerable number of children with None declared.
late CUS or MRI findings did not have significant adverse neurodeve-
lopmental outcomes [33]. Funding sources
Another recent observational study of preterm infants found sig-
nificant associations between cerebral white and gray-matter abnorm- None.
alities on MRI at term equivalent age and risk of adverse neurodeve-
lopmental outcomes at two years of age. Moderate-to-severe cerebral References
white-matter abnormalities on MRI at term equivalent age were pre-
dictive of severe cognitive delay, severe psychomotor delay, cerebral [1] Rees CM, Pierro A, Eaton S. Neurodevelopmental outcomes of neonates with
palsy, and neurosensory impairment. Gray-matter abnormalities were medically and surgically treated necrotizing enterocolitis. Arch Dis Child Fetal
Neonatal Ed 2007;92:F193–8.
also associated with increased risk of cognitive delay, psychomotor [2] Soraisham AS, Amin HJ, Al-Hindi MY, Singhal N, Sauve RS. Does necrotising en-
delay, and cerebral palsy, but less so compared to white-matter ab- terocolitis impact the neurodevelopmental and growth outcomes in preterm infants
normalities. Risk predictions from these MRI findings were independent with birthweight < 1250 g? J Paediatr Child Health 2006;42:499–504.
[3] Dilli D, Eras Z, Ulu HÖ, Dilmen U, Şakrucu ED. Does necrotizing enterocolitis affect
of other perinatal factors and abnormalities on cranial ultrasonography growth and neurodevelopmental outcome in very low birth weight infants? Pediatr
[49]. Surg Int 2012;28:471–6.
The usefulness of MRI as a diagnostic tool to assess brain develop- [4] Sonntag J, Grimmer I, Scholz T, Metze B, Wit J, Obladen M. Growth and neuro-
developmental outcome of very low birthweight infants with necrotizing
ment and injury, either alone or in addition to information available
431
enterocolitis. Acta Paediatr 2000;89:528–32. after intensive neonatal nutritional support. Pediatrics 2009;123:e101–9.
[5] Salhab WA, Perlman JM, Silver L, Sue Broyles R. Necrotizing enterocolitis and [27] Fuglestad A, Rao R, Georgieff MK. The role of nutrition in cognitive development.
neurodevelopmental outcome in extremely low birth weight infants < 1000 g. Handbook of developmental cognitive neuroscience. second ed.Cambridge, MA:
J Perinatol 2004;24:534–40. MIT Press; 2008. p. 623–41.
[6] Schulzke SM, Deshpande GC, Patole SK. Neurodevelopmental outcomes of very low- [28] Gardelis J, Hatzis T, Stamogiannou L, et al. Acitivity of the growth hormone/in-
birth-weight infants with necrotizing enterocolitis. Arch Pediatr Adolesc Med sulin-like growth factor-I axis in critically ill children. J Pediatr Endocrinol
2007;161:583. 2005;18:363–72.
[7] Shah TA, Meinzen-Derr J, Gratton T, et al. Hospital and neurodevelopmental out- [29] Gelato M. The growth hormone/insulin-like growth factor axis in critical illness.
comes of extremely low-birth-weight infants with necrotizing enterocolitis and J Pediatr Endocrinol Metab 2000;13:1023–9.
spontaneous intestinal perforation. J Perinatol 2012;32:552–8. [30] Dammann O, O'Shea M. Cytokines and perinatal brain damage. Clin Perinatol
[8] Hintz SR, Kendrick DE, Stoll BJ, et al. Neurodevelopmental and growth outcomes of 2008;35:643–.
extremely low birth weight infants after necrotizing enterocolitis. Pediatrics [31] Lodha A, Asztalos E, Moore A. Cytokine levels in neonatal necrotizing enterocolitis
2005;115:696–703. and long-term growth and neurodevelopment. Acta Paediatr 2010;99:338–43.
[9] Stoll BJ. Neurodevelopmental and growth impairment among extremely low-birth- [32] Wu YW, Colford JM. Chorioamnionitis as a risk factor. J Am Med Assoc
weight infants with neonatal infection. J Am Med Assoc 2004;292:2357–65. 2000;284:1417–24.
[10] Martin CR, Dammann O, Allred EN, et al. Neurodevelopment of extremely preterm [33] Hintz S, Barnes P, Bulas D, et al. Neuroimaging and neurodevelopmental outcome
infants who had necrotizing enterocolitis with or without late bacteremia. J Pediatr in extremely preterm infants. Pediatrics 2015;135:e32–42.
2010;157:751–756.e1. [34] Volpe J, Kinney H, Rosenberg P. The developing oligodendrocyte: key cellular
[11] Cusick S, Georgieff M. The role of nutrition in brain development: the golden op- target in brain injury in the premature infant. Int J Dev Neurosci 2011;29:565–82.
portunity of the “first 1000 days. J Pediatr Psychol 2016;175:16–21. [35] Volpe JJ. Postnatal sepsis, necrotizing entercolitis, and the critical role of systemic
[12] Vohr BR, Allen M. Extreme prematurity – the continuing dilemma. N Engl J Med inflammation in white matter injury in premature infants. J Pediatr
2005;3521:71–2. 2008;153:160–3.
[13] Ramel SE, Brown LD, Georgieff MK. The impact of neonatal illness on nutritional [36] Shah D, Doyle L, Anderson P, et al. Adverse neurodevelopment in preterm infants
requirements: one size does not fit all. Curr Pediatr Rep 2014;2:248–54. with postnatal sepsis or necrotizing enterocolitis is mediated by white matter ab-
[14] Pfister KM, Gray HL, Miller NC, Demerath EW, Georgieff MK, Ramel SE. Exploratory normalities on magnetic resonance imaging at term. J Pediatr
study of the relationship of fat-free mass to speed of brain processing in preterm 2008;153:170–175.e1.
infants. Pediatr Res 2013;74:576–83. [37] Diaz Heijtz R. Fetal, neonatal, and infant microbiome: perturbations and subsequent
[15] Hansen-Pupp I, Hövel H, Löfqvist C, et al. Circulatory insulin-like growth factor-I effects on brain development and behavior. Semin Fetal Neonatal Med
and brain volumes in relation to neurodevelopmental outcome in very preterm 2016;21:410–7.
infants. Pediatr Res 2013;74:564–9. [38] Penders J, Thijs C, Vink C, et al. Factors influencing the composition of the in-
[16] Lin GC, Robinson DT, Olsen S, et al. Nutritional practices and growth in premature testinal microbiota in early infancy. Pediatrics 2006;118:511–21.
infants after surgical necrotizing enterocolitis. J Pediatr Gastroenterol Nutr [39] Fouhy F, Guinane CM, Hussey S, et al. High-throughput sequencing reveals the
2017;65:111–6. incomplete, short-term recovery of infant gut microbiota following parenteral an-
[17] Thureen PJ, Melara D, Fennessey PV, Hay WW. Effect of low versus high in- tibiotic treatment with ampicillin and gentamicin. Antimicrob Agents Chemother
travenous amino acid intake on very low birth weight infants in the early neonatal 2012;56:5811–20.
period. Pediatr Res 2003;53:24–32. [40] Van Best N, Hornef MW, Savelkoul PHM, Penders J. On the origin of species: factors
[18] Poindexter B, Langer J, Dusick A, Ehrenkranz R. NICHD Neonatal Research shaping the establishment of infant's gut microbiota. Birth Defects Res C Embryo
Network. Early provision of parenteral amino acids in extremely low birth weight Today Rev 2015;105:240–51.
infants: relation to growth and neurodevelopmental outcome. J Pediatr Psychol [41] Mueller NT, Bakacs E, Combellick J, Grigoryan Z, Dominguez-Bello MG. The infant
2006;148:301–5. microbiome development: mom matters. Trends Mol Med 2015;21:109–17.
[19] Stoll BJ, Hansen NI, Bell EF, et al. Neonatal outcomes of extremely preterm infants [42] Douglas-Escobar M, Elliott E, Neu J. Effect of intestinal microbial ecology on the
from the NICHD. Pediatrics 2010;126:443–56. developing brain. JAMA Pediatr 2013;167:374–9.
[20] Pfister KM, Ramel SE. Linear growth and neurodevelopmental outcomes. Clin [43] Blakely ML. Laparotomy versus peritoneal drainage for necrotizing enterocolitis or
Perinatol 2014;41:309–21. isolated intestinal perforation in extremely low birth weight infants: outcomes
[21] Ramel S, Demerath E, Gray H, Younge N, Boys C, Georgieff MK. The relationship of through 18 months adjusted age. Pediatrics 2006;117:e680–7.
poor linear growth velocity with neonatal illness and two year neurodevelopment in [44] Ward CG, Loepke AW. Anesthetics and sedatives: toxic or protective for the de-
preterm infants. Neonatology 2012;102:19–24. veloping brain? Pharmacol Res 2012;65:271–4.
[22] Belfort MB, Rifas-Shiman SL, Sullivan T, et al. Infant growth before and after term: [45] Hack M, Taylor G, Drotar D, et al. Poor predictive validity of the Bayley Scales of
effects on neurodevelopment in preterm infants. Pediatrics 2011;128:e899–906. Infant Development for cognitive function of extremely low birth weight children at
[23] Johnson MJ, Wootton SA, Leaf AA, Jackson AA. Preterm birth and body composi- school age. Pediatrics 2005;116:333–41.
tion at term equivalent age: a systematic review and meta-analysis. Pediatrics [46] Anderson PJ. Underestimation of developmental delay by the new Bayley-III scale.
2012;130:e640–9. Arch Pediatr Adolesc Med 2010;164:352.
[24] Ramel SE, Gray HL, Christiansen E, Boys C, Georgieff MK, Demerath EW. Greater [47] Maalouf E, Duggan P, Counsell S, et al. Comparison of findings on cranial ultra-
early gains in fat-free mass, but not fat mass, are associated with improved neu- sound and magnetic resonance imaging in preterm infants. Pediatrics
rodevelopment at 1 year corrected age for prematurity in very low birth weight 2001;107:719–27.
preterm infants. J Pediatr 2016;173:108–15. [48] Inder T, Wells S, Mogridge N, Spencer C, Volpe J. Defining the nature of the cerebral
[25] Ehrenkranz RA. Growth in the neonatal intensive care unit influences neurodeve- abnormalities in the premature infant: a qualitative magnetic resonance imaging
lopmental and growth outcomes of extremely low birth weight infants. Pediatrics study. J Pediatr 2003;3476:171–9.
2006;117:1253–61. [49] Woodward LJ, Anderson PJ, Austin NC, Howard K, Inder TE. Neonatal MRI to
[26] Franz AR, Pohlandt F, Bode H, et al. Intrauterine, early neonatal, and postdischarge predict neurodevelopmental outcomes in preterm infants. N Engl J Med
growth and neurodevelopmental outcome at 5.4 years in extremely preterm infants 2006;355:685–94.
432

Enterocolitis Necrotizante 2018

Uploaded by

Copyright:

Available Formats

Enterocolitis Necrotizante 2018

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Enterocolitis Necrotizante 2018

Uploaded by

Copyright:

Available Formats

Seminars in Fetal and Neonatal Medicine 23 (2018) 369

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

1744-165X/ © 2018 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

Necrotizing enterocolitis comes in diﬀerent forms: Historical perspectives T

1744-165X/ © 2018 Published by Elsevier Ltd.

“Stage 1, 2 and 3” NEC

uncommon trigger). If the baby is formula-fed, a switch to a hy- 6. Transfusion-related NEC

Conﬂicts of interest presenting as recurrent necrotizing enterocolitis in preterm neonates. J Perinatol

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

A critical analysis of risk factors for necrotizing enterocolitis T

1. Introduction (RCTs) and multicenter or regional/national observational studies

1744-165X/ © 2018 Elsevier Ltd. All rights reserved.

Table 1 0.93–4.79; three studies, 395 infants) [11].

Assessment of risk factors based on studies identiﬁed in search strategy. Please

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

An update on biomarkers of necrotizing enterocolitis T

1. Introduction more on (i) understanding of the pathophysiological mechanisms of

1744-165X/ © 2018 Elsevier Ltd. All rights reserved.

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

Genetic predisposition to necrotizing enterocolitis in premature infants: T

ARTICLE INFO ABSTRACT

1. Introduction technological advances in genotyping platforms, are beginning to yield

1744-165X/ © 2018 Elsevier Ltd. All rights reserved.

Table 1 Table 1 (continued)

2.5. Mannose-binding lectin (MBL) 2.7. Nuclear factor-kappa beta (NFKB)

Trezl et al. [18] Hungary 46/90 Included Bell stage I No No

4. Challenges and limitations 5. Future considerations

polymorphisms with necrotizing enterocolitis in preterm infants. J Pediatr Perinatol 2015;32:905–9.

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

Innate and adaptive immunity in necrotizing enterocolitis T

1744-165X/ © 2018 Elsevier Ltd. All rights reserved.

4. Conclusions function and antigen uptake. Microb Infect 2005;7:997–1004.

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

Necrotizing enterocolitis: The intestinal microbiome, metabolome and T

1744-165X/ © 2018 Elsevier Ltd. All rights reserved.

bacterial species, less diversity and increased proportion of potential

2.3. The intestinal microbial milieu prior to the diagnosis of NEC

Early studies utilizing 16S rDNA for microbiome evaluation sug-

Fig. 2. Microbial proﬁles in necrotizing enterocolitis (NEC).

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

Enteral feeding composition and necrotizing enterocolitis T

1744-165X/ © 2018 Elsevier Ltd. All rights reserved.

Immunoglobulins • Trials used IgG, IgG/IgA and IgG/IgA/IgM

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

The cost of necrotizing enterocolitis in premature infants T

1. Introduction 2. Overall cost and length of stay for birth hospitalization

1744-165X/ © 2018 Published by Elsevier Ltd.

Contents lists available at ScienceDirect

Seminars in Fetal and Neonatal Medicine

Surgical considerations for neonates with necrotizing enterocolitis T