Accepted Manuscript

Interleukin-6 serum levels predict surgical intervention in infants

with necrotizing enterocolitis

Lukas Wisgrill, Anja Weinhandl, Lukas Unterasinger, Gabriele

Amann, Rudolf Oehler, Martin L. Metzelder, Angelika Berger,
Thomas M. Benkoe

PII: S0022-3468(18)30490-1
DOI: doi:10.1016/j.jpedsurg.2018.08.003
Reference: YJPSU 58768
To appear in: Journal of Pediatric Surgery
Received date: 20 January 2018
Revised date: 13 June 2018
Accepted date: 2 August 2018

Please cite this article as: Lukas Wisgrill, Anja Weinhandl, Lukas Unterasinger, Gabriele
Amann, Rudolf Oehler, Martin L. Metzelder, Angelika Berger, Thomas M. Benkoe ,
Interleukin-6 serum levels predict surgical intervention in infants with necrotizing
enterocolitis. Yjpsu (2018), doi:10.1016/j.jpedsurg.2018.08.003

This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.
 ACCEPTED MANUSCRIPT

Interleukin-6 serum levels predict surgical intervention in infants with

necrotizing enterocolitis

Lukas Wisgrill1, Anja Weinhandl2, Lukas Unterasinger1, Gabriele Amann3, Rudolf Oehler4,

Martin L. Metzelder2, Angelika Berger1, Thomas M. Benkoe2

PT
1
Department of Pediatrics and Adolescent Medicine, Division of Neonatology, Pediatric

RI
Intensive Care & Neuropediatrics, Medical University of Vienna, Vienna, Austria
2

SC
Department of Pediatric Surgery, Medical University of Vienna, Vienna, Austria
3
Institute of Clinical Pathology, Medical University of Vienna, Vienna, Austria
4
NU
Surgical Research Laboratories, Department of Surgery and Comprehensive Cancer Center,
Medical University of Vienna, Vienna, Austria
MA

Running headline: IL-6 in necrotizing enterocolitis

ED

Correspondence:
PT

Thomas M. Benkoe, M.D.

Department of Pediatric Surgery

CE

Medical University of Vienna

Währinger Gürtel 18-20, 1090 Vienna, Austria

AC

eMail: thomas.benkoe@meduniwien.ac.at

Phone: +43 1 40400 19254

FAX: +43 1 40400 68380

Funding: This research did not receive any specific grant from funding agencies in the

public, commercial, or not-for-profit sectors.

1
 ACCEPTED MANUSCRIPT

Abstract

Background: Symptoms at suspicion of necrotizing enterocolitis (NEC) are often non-

specific and several biomarkers have been evaluated for their discriminative power to both

diagnose and predict the course from NEC suspicion to complicated disease requiring surgical

intervention. Thus, we aimed to assess the utility of interleukin-6 (IL-6) to predict surgical

PT
intervention in infants suffering from NEC and, furthermore, to discriminate infants with

RI
starting NEC or late-onset sepsis (LOS).

SC
Methods: IL-6 serum levels at disease onset were retrospectively analyzed in 24 infants

suffering from NEC as well as 16 neonates with LOS.

NU
Results: IL-6 serum levels at disease onset were significantly higher in infants suffering from

NEC necessitating surgical intervention in the disease course compared to infants with
MA

medical NEC (5000 [785-5000] vs. 370 [78-4716] pg/ml, p=0.0008) as well as gram-positive

LOS (5000 [785-5000] vs. 84 [12-269] pg/ml, p=0.0001). Infants suffering from gram-
ED

negative LOS exhibited elevated IL-6 serum levels at disease onset comparable to infants

with surgical NEC (5000 [1919-5000] vs. 5000 [785-5000] pg/ml, p =1.00).
PT

Conclusion: The pro-inflammatory cytokine IL-6 appears to be a promising marker to

CE

distinguish surgical NEC from medical NEC at the onset of disease but cannot discriminate

between surgical NEC and gram-negative LOS.

AC

Keywords: necrotizing enterocolitis, interleukin 6, biomarker, premature infants

Level of Evidence: II

Abbreviations: NEC - necrotizing enterocolitis; IL – interleukin; LOS – late-onset sepsis;

GN - gram-negative; GP - gram-positive

2
 ACCEPTED MANUSCRIPT

1 Introduction

Necrotizing enterocolitis (NEC) is a life-threatening gastrointestinal complication in

premature infants and is associated with high morbidity and mortality, leading to long-term

sequelae in surviving neonates [1, 2]. Despite different preventive approaches, the number of

deaths related to NEC among premature neonates is increasing [3-5].

NEC pathogenesis is not completely understood, but is believed to involve a dysregulated

PT
excessive inflammatory response of the immature intestinal epithelium of neonates to luminal

RI
bacteria, leading to exaggerated inflammatory damage of the epithelial gut barrier.

SC
Additionally, imbalance of intestinal microperfusion as well as intestinal ischemia might

provide further damage to the gut epithelium, leading to a “leaky gut” with increased bacterial
NU
translocation and, subsequently, a systemic inflammatory immune response [4, 6].

Plasma cytokine levels are very attractive as diagnostic tools in neonates because they are
MA

easily accessible and can be processed in a standardized fashion requiring only small blood

volumes. Recent data suggest that only few cytokines are significantly influenced by the
ED

presence of NEC, two of the most relevant pro-inflammatory cytokines being interleukin (IL)-
PT

8 and IL-6 [7-9]. Previous studies from our institution on IL-8 levels and NEC indicated a

significant correlation with disease extent, need for surgery, and mortality [10-12]. In contrast
CE

to IL-8, clinical data on IL-6 and NEC is sparse. Although there is direct evidence that the

onset of NEC results in significantly elevated IL-6 levels, no studies thus-far have
AC

investigated the potential value of IL-6 to discriminate NEC from LOS as well as to predict

the disease course during NEC episodes [7, 8].

We hypothesized that IL-6 levels are significantly elevated during the initial phase of NEC

and that the levels of IL-6 correlate with the subsequent need for surgical intervention.

Secondary aims of the study were to analyse IL-6 levels in LOS according to the presence of

gram-positive and gram-negative bacteria and to evaluate the ability of IL-6 levels to

distinguish, at disease onset, between sepsis and NEC.

3
 ACCEPTED MANUSCRIPT

2 Methods

2.1 Study design

The study was conducted at the Division of Neonatology and the Department of Pediatric

Surgery at the Medical University of Vienna and was approved by the local ethics committee

(EC no. 1374/2017). Neonates with diagnosed NEC and blood-culture positive late-onset

sepsis were recruited during October 2015 and January 2017. Infants exhibiting clinical signs

PT
and radiographic findings of NEC were staged according to the modified Bell’s staging

RI
criteria [13]. Depending on the subsequent clinical course, infants with NEC were divided

SC
into two different treatment groups, medical and surgical NEC. Infants assigned to the

medical group had no surgical intervention during the disease course. Indications for surgery
NU
included evidence of intestinal perforation and/or clinical deterioration despite maximum

conservative treatment [14]. Decision for surgical intervention was dictated by the clinical
MA

judgement of the attending pediatric surgeon in coordination with the attending neonatologist

and was in no way influenced by this study. The diagnosis of NEC was confirmed intra-
ED

operatively in all patients. Macroscopic necrotic tissue was resected saving as much as
PT

possible viable tissue. The diagnosis of NEC was confirmed by histopathology in all patients

who received bowel resections. According to macroscopically evident affected bowel, disease
CE

extent was defined as focal NEC when it was limited to a well defined single intestinal

segment up to 20cm of length, multifocal NEC if it included 2 or more non-continuous

AC

intestinal segments and panintestinal NEC when the majority of the small and/or large

intestines were involved. Infants with isolated intestinal perforation were excluded from the

study. Since 2015, IL-6 is part of the routine sepsis surveillance protocol at our department.

Blood samples were obtained from deteriorating infants at the onset of clinical signs of sepsis

or NEC. To compare IL-6 values at the time of NEC diagnosis with preoperative IL-6 levels,

we additionally evaluated preoperative IL-6 levels in the group of surgical NEC, which were

obtained within 6 hours prior surgery. Diagnosis of culture-proven bloodstream infections

4
 ACCEPTED MANUSCRIPT

was based on NeoKISS definitions: Culture-proven blood stream infections required the

recognition of a pathogen from one or more blood cultures, not related to another infection

site, and at least 2 of the following clinical signs: temperature >38°C or <36.5°C or

temperature instability, tachycardia or bradycardia, apnea, prolonged capillary refill time,

metabolic acidosis, hyperglycemia or other sepsis signs, such as increased oxygen

requirement or apathy. Diagnosis of blood stream infections with coagulase-negative

PT
staphylococci (CoNS) additionally required at least one of the following positive laboratory

RI
parameters of infection: C-reactive protein > 2 mg/dl or IL-6 > 50 pg/ml, immature/total

SC
neutrophil ratio > 0.2, white blood cell (WBC) count < 5000/µL and platelets <100/nl.

2.2 Data collection

NU
Demographic, clinical and laboratory data were routinely collected during the hospital stay
MA

using the patient data management system ICCA of Phillips Healthcare. Blood culture results

and antibiotic resistance profiles were analysed using the monitoring of microorganism
ED

database of Medexter Healthcare [15].

2.3 Determination of IL-6

PT

IL-6 serum concentrations were routinely tested using the IL-6 electrochemiluminescent
CE

immunoassay (ECLIA) using a Cobas e411analyzer (Roche, Mannheim, Germany). In our

assay, IL-6 values less than 50 pg/ml are considered normal for infants [16]. The assay was
AC

calibrated up to 5000 pg/ml. At this level the normal value was exceeded by more than 100-

fold and therefore not further diluted to determine the exact amount of IL-6.

2.4 Statistical analysis

Patient characteristics are presented as mean ± standard deviation or as frequency within the

study group if not stated otherwise. The Kruskal-Wallis-Test with post-hoc Dunn-Bonferroni

correction was used to test statistical significance. ROC analysis was performed with SPSS. A

5
 ACCEPTED MANUSCRIPT

p-value <0.05 was considered statistically significant. Statistical analysis was performed

using SPSS (IBM, Armonk, USA).

3 Results

3.1 Patient characteristics

PT
During the 16-month study period, a total of 24 neonates with NEC were included as well as

RI
16 control infants with diagnosis of culture-proven LOS. Clinical characteristics of the

SC
enrolled infants are summarized in table 1. Both groups exhibited similar gestational age and

age at diagnosis of infection. Infants in the sepsis group display significantly lower birth
NU
weights compared to infants with NEC (p = 0.044). In the sepsis group, gram-negative late
MA

onset sepsis (GN-LOS) was observed in 8 infants, and gram-positive LOS (GP-LOS) was

detected in 8 cases.
ED

In the NEC group, surgical treatment was performed in 12 infants (50%), and medical

treatment in 12 cases (50%). Of 24 infants with NEC, six exhibited Bell Stage I, six Bell stage
PT

II and 12 patients displayed Bell stage III. The median time from NEC diagnosis to surgery

was 24 [8 - 96] hours. Indications for surgery were radiographic signs of intestinal perforation
CE

in 9, clinical signs of perforation (e.g. palpable resistance, thickening and discoloured

AC

abdominal wall) in 2 and deterioration despite maximum treatment in one infant.

Intraoperative findings and surgical procedures are summarized in table 2.

Of the operated infants, none had focal NEC (0%), 9 infants had multifocal NEC (75%) and 3

infants had panintestinal NEC (25%). Intestinal perforations were present in 11 infants with

surgical NEC. Intestinal resections were done in 9 patients, 4 infants were treated with small

bowel and 5 infants with small and large bowel resections. Two infants presented with

characteristic morphologic changes of NEC (e.g. mosaic like pattern of affected bowel and

6
 ACCEPTED MANUSCRIPT

pneumatosis intestinalis) and intestinal perforations in an otherwise non-necrotic bowel

segment. These infants were treated with a diverting enterostomy and the intestinal

perforation was sutured but not resected. One infant was treated with a diverting enterostomy

proximal the affected and non-necrotic bowel.

3.2 IL-6 is significantly higher in surgical NEC patients compared to medical NEC patients

PT
RI
Of all evaluated laboratory parameters, only serum concentrations of IL-6 were significantly

SC
higher in surgical NEC compared to cases with medical NEC (5000 [785-5000] vs. 370 [78-

4716], p = 0.008). In 8 of 12 cases with surgical NEC, IL-6 serum levels exceeded the assay
NU
range of 5000 pg/ml (Table 3/ Figure 1). The comparison of IL-6 levels at disease onset with

preoperative IL-6 values did not reach a statistical significance (p=0.11). IL-6 values tended
MA

to remain constant or slightly decrease between the time of NEC diagnosis and the time of

operation (Table 2). An explorative ROC analysis was performed to discriminate surgical
ED

NEC from medical NEC based on the IL-6 level at the time of NEC diagnosis. The AUC was

0.931 (CI 0.83 – 1.0), and an exploratory cut-off point could be established at 1440 pg/ml. At
PT

this value IL-6 reached a sensitivity of 91.7% and a specificity of 83.3% (Figure 2). IL-6
CE

levels were elevated according to the infants Bell´s stage. IL-6 levels showed a stage-

depending increase of IL-6 levels with higher levels in Bell stage III compared to lower levels
AC

in Bell stage II and I (5000 [785-5000] vs. 596 [152-4716] vs. 158 [78-2578] pg/ml,

p=0.001). In addition, we excluded Bell stage I cases in the medical NEC group and found no

difference in the AUC (surgical NEC vs. medical NEC without Bell stage I AUC 0.931; CI

0.81 – 1.0; data not shown).

Furthermore, we investigated if NEC infants with proven blood stream infection showed

different IL-6 levels when compared to NEC cases with negative blood culture results. In the

7
 ACCEPTED MANUSCRIPT

surgical NEC group, 6 of 12 infants showed positive blood cultures (4 E. coli and 2 K.

pneumoniae) compared to 3 of 12 infants in the medical NEC group (2 K. pneumoniae and 1

E. faecalis). In the surgical group, positive blood cultures were not associated with altered IL-

6 serum levels (5000 [785-5000] vs. 3402 [1443-5000] pg/ml, p = 0.36). In contrast, medical

NEC infants with positive blood cultures showed a trend towards decreased IL-6 serum levels

(152 [102-1438] vs. 543 [78-4716] pg/ml, p = 0.33).

PT
RI
3.3 IL-6 levels distinguish GN-LOS from GP-LOS and medical NEC, but not from surgical

SC
NEC
NU
IL-6 is an established early biomarker for neonatal sepsis. Therefore, we included infants with

culture-proven GN-LOS (6 E. coli and 2 K. pneumoniae) and GP-LOS (6 CoNS and 2 S.

MA

aureus). Infants suffering GN-LOS showed significantly higher serum concentrations of IL-6

compared to neonates with medical NEC (5000 [1919-5000] vs. 370 [78-4716] pg/ml, p =
ED

0.014) and GP-LOS (5000 [1919-5000] vs. 84 [12-269] pg/ml, p = 0.0001). IL-6 serum levels

did not differ in patients with surgical NEC and GN-LOS (5000 [785-5000] vs. 5000 [1919-
PT

5000] pg/ml, p = 1.00). Interestingly, NEC infants treated medically displayed higher
CE

thrombocyte counts compared to the other study groups, reaching borderline statistical

significance compared to the GP-LOS group (260 [42-385] vs. 98 [60-211] x 103/µl, p =
AC

0.048). WBC and CRP levels did not differ among the study groups. IL-6 serum levels

showed no correlation with other laboratory values (data not shown). However, IL-6 serum

levels could not discriminate between surgical NEC and GN-LOS (AUC 0.458; 0.19 – 0.71).

Data are depicted in Figure 2.

4 Discussion

8
 ACCEPTED MANUSCRIPT

The present study demonstrates that IL-6 plasma levels strongly correlate with disease

severity and the need for surgical intervention in neonates with NEC. At disease onset IL-6

plasma levels are significantly higher in neonates with surgical NEC compared to patients

with medically treated NEC. The observed IL-6 levels further correlate with advanced Bell

stages in infants suffering from NEC. The present data supports our hypothesis that elevated

IL-6 levels correspond with the disease severity of NEC. It further highlights the importance

PT
of IL-6 as a promising diagnostic marker in neonates with NEC.

RI
The cytokine response at the onset of NEC has received much scientific attention [17, 18].

SC
This has revealed an extensive dysregulation of gene expression in human NEC tissues with

significant changes in the upregulation of proinflammatory cytokines [19]. Several studies

NU
have identified IL-6 and IL-8 as key proinflammatory cytokines in NEC [7, 9, 17]. Most

recently, we were able to elucidate the role of IL-8 in NEC patients and demonstrate that IL-8
MA

levels predicted the extent of involved intestine [10], the need for surgical intervention, and

60-day mortality [12]. Routine IL-8 measurements were exchanged for IL-6 due to a
ED

rearrangement of the laboratory core facilities at our institution. Unfortunately, we are

therefore unable to provide information on both IL-6 and IL-8 levels in the present patient
PT

cohort. Based on our previous experience with IL-8 and the valuable information it provided
CE

in the risk assessment in infants with NEC, we found it promising to evaluate IL-6 as a

diagnostic biomarker in infants with NEC.

AC

The diagnosis of NEC currently relies on clinical symptoms and radiological features such as

pneumatosis intestinalis and pneumoperitoneum, underlying the Bell´s stages of disease [13].

Initial clinical signs are often unspecific and mimic feeding intolerance or other forms of

benign gastroenteritis, which hampers early diagnosis [4]. However, the typical clinical

scenario of a premature infant who is at first developing subtle signs of feeding intolerance

and then progressing to abdominal distention and bloody stool, often with sepsis and

9
 ACCEPTED MANUSCRIPT

hemodynamic instability, is frequently abbreviated by those who deteriorate rapidly within a

few hours, presenting with pan-intestinal disease extent. It is important to note that early

clinical signs in the initial phase of both NEC and sepsis are overlapping and thus difficult to

discriminate. This early phase of the disease with unspecific clinical signs is often referred as

Bell Stage I without common radiological findings such as pneumatosis or aeroportogram.

Despite all limitations of the Bell Staging criteria, infants with NEC Stage I were included in

PT
the study to determine the utility of IL-6 as an early diagnostic marker. The early recognition

RI
or exclusion of NEC, even in the stated Bell Stage I, might have important impacts on the

SC
prospective clinical management of those neonates (e.g. stop enteral feeding, radiographic

monitoring, change of antibiosis, early transfer into a tertiary care center). This circumstance
NU
mandates future research focused on implementing targeted therapy at an early stage.

Data on the predictive value of IL-6 in neonates with NEC are promising but sparse. The
MA

presented results are in line with a prior study on plasma concentrations of 6 different

cytokines that reported a significant increase of IL-6, IL-8, TNF and IL-1ß levels in surgical
ED

NEC compared to spontaneous intestinal perforation and controls [8]. Despite concomitant

up-regulation of IL-6 and IL-8, only the latter was able to discriminate between limited NEC
PT

and NEC totalis in that study. In contrast to our study, infants with medical NEC and culture-
CE

positive LOS were not considered.

As a consequence of this diagnostic dilemma we further aimed to investigate the utility of IL-
AC

6 as a reliable marker to distinguish patients with NEC or LOS at disease onset. Harris and

colleagues [7] reported that IL-6 levels were 6- to 10-fold elevated in infants with NEC

compared to infants suffering from sepsis and healthy controls. These elevated IL-6 levels

were only observed in NEC infants with positive blood culture results. In their study IL-6

levels were not able to discriminate between sepsis and NEC. However, a further

classification of infections by gram-positive or gram-negative bacteria was not included.

Although we observed significantly higher IL-6 levels in gram-negative LOS compared with

10
 ACCEPTED MANUSCRIPT

gram-positive LOS, IL-6 levels at disease onset were unable to discriminate between infants

with NEC and gram-negative LOS. Enterobacteriaceae appear to be potent activators of

proinflammatory pathways leading to exaggerated IL-6 secretion upon infection.

Lipopolysaccharide (LPS), a major component of the outer membrane of gram-negative

bacteria, was identified as the top upstream transcriptional regulator in NEC. LPS induces IL-

1, IL-6, TNF, and IFN, which in turn activate inflammatory transcription factors [21]. The

PT
identification of LPS as the top upstream regulator is consistent with increasing evidence that

RI
infants who develop NEC display a gut microbial dysbiosis with overrepresentation of gram-

SC
negative bacteria prior to disease onset [22, 23]. This hypothesis is supported by clinical

evidence that oral administration of aminoglycosides appears to be protective against NEC

NU
[24].

The early or potential discrimination of infections with gram-negative or gram-positive

MA

bacteria in infants with sepsis and sepsis syndrome would be highly desirable. In adults,

procalcitonin levels were able to predict gram-positive or gram-negative sepsis, displaying

ED

higher values in gram-negative bacteremia [20]. In the context of adapted empirical antibiotic

schemes IL-6 values could provide valuable additional information at sepsis suspicion.
PT

However, larger studies are needed to assess the potential role of IL-6 to predict gram-
CE

negative bacteraemia in neonatal sepsis.

AC

It is of utmost importance to note that the dominant factor that defines the further course,

subsequent gastrointestinal morbidity, and outcome of infants with NEC is the extent of

affected intestine. The surgical option employed is strongly influenced by clinical judgement

and estimation of the involved intestinal tissue [25]. The evaluation of different surgical

procedures in neonates with NEC in the past three decades has clearly shown that the

modality of surgical treatment is not able to significantly influence the outcome [4]. Surgical

intervention is associated with a variety of inherent stresses that may aggravate the

11
 ACCEPTED MANUSCRIPT

physiologic derangement observed in NEC [26]. One of the key questions that remains to be

answered is the timing of surgery. While the traditional concept to proceed to surgery in the

absence of clinical or radiological signs of intestinal perforation when there is a worsening of

the clinical status of the patient despite optimal treatment, the additional information provided

by biomarkers such as IL6 might justify re-assessment of current surgical algorithms. It might

be prudent to undergo surgery earlier in patients with excessively high cytokine levels, at a

PT
time point when these fragile but otherwise stable infants might be able to tolerate surgery

RI
better than at a later more advanced disease stage.

SC
A limitation of our study is the upper detection limit of IL-6, which represents a 100-fold

increase of its normal value. The upper limit was reached in 8 out of 12 cases with surgical
NU
NEC and in 5 out of 8 cases with gram-negative LOS, which limits further evaluation. Since

the patients sera were not further diluted to measure the exact quantity, we are not able to
MA

assess potential differences in elevated IL-6 serum concentrations in those two groups. As a

further consequence of the excessively high IL-6 levels observed in surgical NEC we are not
ED

able to correlate IL-6 levels with intraoperative findings. As a consequence of those results,

samples for IL-6 measurement are now being diluted to measure the exact serum
PT

concentrations of IL-6 in neonates at our institution.

CE

In conclusion, the proinflammatory cytokine IL-6 was able to predict the subsequent need for
AC

surgical intervention at the disease onset in our patient cohort. It is important to note that the

observed IL-6 levels remained highly elevated between NEC onset and operation. This further

underlines the diagnostic value of IL-6 to predict the need for surgical intervention at disease

onset. Whether an early surgical intervention in the presence of extremely high cytokine

levels, even in the absence of evidence for intestinal perforation, can improve the outcome of

this devastating disease needs to be tested in prospective treatment trials. Further studies are

12
 ACCEPTED MANUSCRIPT

also needed to assess the association of IL-6 serum levels with the extent of affected intestine

in surgical NEC cases as well as mortality in infants suffering from NEC.

Acknowledgments: We thank David Berry for critically reviewing the manuscript.

Conflict of interest: The authors declare no potential conflict of interest.

Funding: This research did not receive any specific grant from funding agencies in the

PT
public, commercial, or not-for-profit sectors.

RI
SC
NU
MA
ED
PT
CE
AC

13
 ACCEPTED MANUSCRIPT

References

[1] Hintz SR, Kendrick DE, Stoll BJ, Vohr BR, Fanaroff AA, Donovan EF, et al.

Neurodevelopmental and growth outcomes of extremely low birth weight infants

after necrotizing enterocolitis. Pediatrics 2005;115(3):696-703.

[2] Lin PW, Stoll BJ. Necrotising enterocolitis. Lancet 2006;368(9543):1271-83.

PT
[3] Patel RM, Myers LS, Kurundkar AR, Maheshwari A, Nusrat A, Lin PW. Probiotic

RI
bacteria induce maturation of intestinal claudin 3 expression and barrier function.

SC
Am J Pathol 2012;180(2):626-35.

[4] Neu J, Walker WA. Necrotizing enterocolitis. The New England journal of medicine
NU
2011;364(3):255-64.
MA

[5] AlFaleh K, Anabrees J. Probiotics for prevention of necrotizing enterocolitis in

preterm infants. Evid Based Child Health 2014;9(3):584-671.

ED

[6] Markel TA, Crisostomo PR, Wairiuko GM, Pitcher J, Tsai BM, Meldrum DR. Cytokines

in necrotizing enterocolitis. Shock 2006;25(4):329-37.

PT

[7] Harris MC, Costarino AT, Jr., Sullivan JS, Dulkerian S, McCawley L, Corcoran L, et al.
CE

Cytokine elevations in critically ill infants with sepsis and necrotizing enterocolitis. J

Pediatr 1994;124(1):105-11.
AC

[8] Bhatia AM, Stoll BJ, Cismowski MJ, Hamrick SE. Cytokine levels in the preterm infant

with neonatal intestinal injury. Am J Perinatol 2014;31(6):489-96.

[9] Benkoe T, Baumann S, Weninger M, Pones M, Reck C, Rebhandl W, et al.

Comprehensive evaluation of 11 cytokines in premature infants with surgical

necrotizing enterocolitis. PloS one 2013;8(3):e58720.

14
 ACCEPTED MANUSCRIPT

[10] Benkoe T, Reck C, Gleiss A, Kettner S, Repa A, Horcher E, et al. Interleukin 8 correlates

with intestinal involvement in surgically treated infants with necrotizing enterocolitis.

J Pediatr Surg 2012;47(8):1548-54.

[11] Benkoe TM, Mechtler TP, Weninger M, Pones M, Rebhandl W, Kasper DC. Serum

levels of interleukin-8 and gut-associated biomarkers in diagnosing necrotizing

PT
enterocolitis in preterm infants. J Pediatr Surg 2014;49(10):1446-51.

[12] Benkoe T, Reck C, Pones M, Weninger M, Gleiss A, Stift A, et al. Interleukin-8 predicts

RI
60-day mortality in premature infants with necrotizing enterocolitis. J Pediatr Surg

SC
2014;49(3):385-9.

[13] Walsh MC, Kliegman RM. Necrotizing enterocolitis: treatment based on staging
NU
criteria. Pediatr Clin North Am 1986;33(1):179-201.
MA

[14] Henry MC, Moss RL. Current issues in the management of necrotizing enterocolitis.

Semin Perinatol 2004;28(3):221-33.

ED

[15] Adlassnig KP, Berger A, Koller W, Blacky A, Mandl H, Unterasinger L, et al. Healthcare-

associated infection surveillance and bedside alerts. Stud Health Technol Inform
PT

2014;198:71-8.
CE

[16] Panero A, Pacifico L, Rossi N, Mancuso G, Stegagno M, Chiesa C. Interleukin 6 in

neonates with early and late onset infection. The Pediatric infectious disease journal
AC

1997;16(4):370-5.

[17] Maheshwari A, Schelonka RL, Dimmitt RA, Carlo WA, Munoz-Hernandez B, Das A, et

al. Cytokines associated with necrotizing enterocolitis in extremely-low-birth-weight

infants. Pediatric research 2014;76(1):100-8.

15
 ACCEPTED MANUSCRIPT

[18] Chan KY, Leung FW, Lam HS, Tam YH, To KF, Cheung HM, et al. Immunoregulatory

protein profiles of necrotizing enterocolitis versus spontaneous intestinal perforation

in preterm infants. PloS one 2012;7(5):e36977.

[19] Chan KY, Leung KT, Tam YH, Lam HS, Cheung HM, Ma TP, et al. Genome-wide

expression profiles of necrotizing enterocolitis versus spontaneous intestinal

PT
perforation in human intestinal tissues: dysregulation of functional pathways. Ann

Surg 2014;260(6):1128-37.

RI
[20] Guo SY, Zhou Y, Hu QF, Yao J, Wang H. Procalcitonin is a marker of gram-negative

SC
bacteremia in patients with sepsis. Am J Med Sci 2015;349(6):499-504.

[21] MohanKumar K, Namachivayam K, Cheng F, Jiang RH, Flores-Torres J, Torres BA, et al.
NU
Trinitrobenzene sulfonic acid-induced intestinal injury in neonatal mice activates
MA

transcriptional networks similar to those seen in human necrotizing enterocolitis.

Pediatric research 2017;81(1-1):99-112.

ED

[22] Warner BB, Deych E, Zhou Y, Hall-Moore C, Weinstock GM, Sodergren E, et al. Gut

bacteria dysbiosis and necrotising enterocolitis in very low birthweight infants: a

PT

prospective case-control study. Lancet 2016;387(10031):1928-36.

CE

[23] Wang Y, Hoenig JD, Malin KJ, Qamar S, Petrof EO, Sun J, et al. 16S rRNA gene-based

analysis of fecal microbiota from preterm infants with and without necrotizing
AC

enterocolitis. ISME J 2009;3(8):944-54.

[24] Bury RG, Tudehope D. Enteral antibiotics for preventing necrotizing enterocolitis in

low birthweight or preterm infants. Cochrane Database Syst Rev 2001(1):CD000405.

[25] Hall NJ, Eaton S, Pierro A. Royal Australasia of Surgeons Guest Lecture. Necrotizing

enterocolitis: prevention, treatment, and outcome. J Pediatr Surg 2013;48(12):2359-

67.

16
 ACCEPTED MANUSCRIPT

[26] Raval MV, Moss RL. Current concepts in the surgical approach to necrotizing

enterocolitis. Pathophysiology 2014;21(1):105-10.

PT
RI
SC
NU
MA
ED
PT
CE
AC

17
 ACCEPTED MANUSCRIPT

Figure Legends

Figure 1 Laboratory parameters in neonates among the study population. White blood cell

count (WBC; A), platelet count (B) as well as serum concentrations of C-reactive protein

(CRP; D) and interleukin 6 (IL-6; E) in surgical NEC (n=12), medical NEC (n=12), gram

negative (GN) sepsis (n=8) and gram positive (GP) sepsis (n=8) patients. IL-6 serum levels in

PT
infants suffering NEC stage I (n=6), stage II (n=6) and stage III (n=12; C). Boxes display the

25th and 75th percentiles and error bars indicate the 5th and the 95th percentiles. Median values

RI
are represented by the box middle line. * = p<0.05; n.s. = not significant

SC
Figure 2 Receiver operating characteristics curve analysis of IL-6 serum levels in infants with
NU
surgical NEC (sNEC) compared to medical NEC (mNEC- solid line), sNEC compared to late

onset sepsis (LOS - large dotted line) and sNEC compared to gram-negative sepsis (GN -
MA

small dotted line) sepsis. The calculated exploratory IL-6 cut-off value between sNEC and

mNEC was 1440 pg/ml with a sensitivity of 91.7% and a specificity of 83.3%. Data are
ED

presented as area under the curve (AUC) with 5 – 95% confidence interval.
PT
CE
AC

18
 ACCEPTED MANUSCRIPT

Table 1. Demographic data of neonates according to study groups

Surgical Medical NEC GN-LOS GP-LOS

NEC (n=12) (n=8) (n=8)
(n=12)
Gestational age, wks 25.9 [23.1 - 27.3 [24.1 – 27.3 24.5
(median [min-max]) 38.4] 33.4] [23.5 – 27.1 [24 – 28.2]

PT
Birth weight, g (median 750 [630 - 902 [590 - 692 [585 – 665 [565 –
[min-max]) 3160] 2110] 1090] 915]

RI
Age at Onset, d (median 14.5 [4 - 28] 13.5 [3 - 34] 15 [5 – 25] 11

SC
[min-max]) [4 – 26]
Mechanical Ventilation, 12/12 (100%) 4/12 (33%) 6/8 (75%) 0/8 (0%)
NU
n (%)
Use of inotropes, n (%) 7/12 (58%) 3/12 (25%) 5/8 (63%) 0/8 (0%)
Twins, n (%) 1/12 (8.3%) 3/12 (25%) 2/8 (25%) 3/8 (37.5%)
MA

Steroids, n (%) 8/12 (66.7%) 8/12 (66.7%) 5/8 (62.5%) 5/8 (62.5%)
IVH, n (%) 3/12 (25%) 1/12 (8.3%) 1/8 (12.5%) 1/8 (12.5%)
ED

BPD, n (%) 2/12 (16.7%) 3/12 (25%) 3/8 (37.5%) 3/8 (37.5%)
ROP, n (%) 7/12 (58.3%) 3/12 (25%) 3/8 (37.5%) 4/8 (50%)
PT

PDA, n (%) 7/12 (58.3%) 6/12 (50%) 8/8 (100%) 7/8 (87.5%)
Medical closure of PDA, 7/12 (58.3%) 6/12 (50%) 8/8 (100%) 7/8 (87.5%)
CE

n (%)
Ligation of PDA, n (%) 1/12 (8.3%) 1/12 (8.3%) 1/8 (12.5%) 0/8 (0%)
AC

60d mortality, n (%) 4/12 (33.3%) 0/12 (0%) 2/8 (25%) 0/8 (0%)
NEC, necrotizing enterocolitis; GN, gram-negative; GP, gram-positive; LOS, late-onset

sepsis; n.a., not applicable; IVH, intraventricular hemorrhage; BPD, bronchopulmonary

dysplasia; ROP, retinopathy of prematurity; PDA, patent ductus arteriosus

19
 ACCEPTED MANUSCRIPT

Table 2. Demographic and laboratory features of infants with surgically and medically

treated NEC

Variable Surgical NEC Medical NEC p-value

Gestational age, wks 25.9 [23.1 - 38.4] 27.3 [24.1 – 33.4] 0.838

PT
Birth weight, g 750 [630 - 3160] 902 [590 - 2110] 0.963

Age at Onset, d 14.5 [4 - 28] 13.5 [3 - 34] 0.737

RI
NEC Bell stage 12/12 (100%) III 6/12 (50%) I n.a.

SC
6/12 (50%) II

Hours until surgery, h 24 [8 - 96] n.a. n.a.

NU
60-d mortality, n (%) 4 (33.3%) 0 (0%) 0.093
MA

IL-6 (pg/ml) at onset 5000 [785-5000] 370 [78-4716] 0.0001

IL-6 (pg/ml) at surgery 3646 [785 - 5000] n.a. n.a.

ED

Perforation, n (%) 11 (91.6%) 0 (0%) 0.0001

Resection, n Small intestine: 4 n.a. n.a

PT

Colon: 0

Both: 5
CE

None: 3
AC

Length of resection (cm) 11 [0 - 40] n.a. n.a.

Data are presented as medians [minimum-maximum] if not stated otherwise.

20
 ACCEPTED MANUSCRIPT

Table 3. Laboratory parameters of neonates according to the study groups

Surgical NEC Medical NEC GN-LOS GP-LOS

(n=12) (n=12) (n=8) (n=8)
3
WBCs (10 /µl) 6.2 [1.8-22.3] 9.9 [4.2-22.5] 5.4 [0.9-30.5] 10.7 [6.2-32.1]
Platelets (103/µl) 192 [22-547] 260 [42-385] 118 [90-277] 98 [60-211]
Monocytes (103/µl) 376 [39-2168] 997 [503-2121] 542 [67-702] 1462 [215-

PT
4711]
CRP (mg/dl) 6.3 [0.2-19.2] 1.7 [0.2-6.5] 3.4 [0.2-6.6] 3.6 [1.0-4.4]

RI
IL-6 (pg/ml) 5000 [785- 370 [78-4716] 5000 [1919- 84 [12-269]

SC
5000]) 5000]
Positive blood 6/12 (50%) 3/12 (25%) 8/8 (100%) 8/8 (100%)
NU
cultures, n (%)
Data shown as median [min-max]; NEC, necrotizing enterocolitis; GN, gram-negative; GP,
MA

gram-positive; LOS, late-onset sepsis; WBC, white blood count; CRP, C-reactive protein; IL,

interleukin
ED
PT
CE
AC

21
Figure 1
Figure 2