Cytokine xxx (2013) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Cytokine

journal homepage: www.journals .e lsev ier .com/cytokine

Neonatal modulation of serum cytokine profiles by a specific mixture ofanti-inflammatory neutral and acidic oligosaccharides in preterm infants

1043-4666/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.cyto.2013.07.002

⇑ Corresponding author. Address: Room KTC 4X033, VU University MedicalCenter, PO Box 7057, 1007 MB Amsterdam, The Netherlands. Tel.: +31 20 444 6206;fax: +31 20 444 2422.

E-mail address: jp.vandenberg@vumc.nl (J.P. van den Berg).

Please cite this article in press as: van den Berg JP et al. Neonatal modulation of serum cytokine profiles by a specific mixture of anti-inflammatoryand acidic oligosaccharides in preterm infants. Cytokine (2013), http://dx.doi.org/10.1016/j.cyto.2013.07.002

Jolice P. van den Berg a,⇑, Ninke van Zwieteren a, Elisabeth A.M. Westerbeek a, Johan Garssen b,c,Ruurd M. van Elburg a,c

a Department of Paediatrics, Division of Neonatology, VU University Medical Center, Amsterdam, The Netherlandsb Division of Pharmacology, Utrecht Institute for Pharmaceutical Sciences, Beta Faculty, Utrecht University, The Netherlandsc Center for Specialised Nutrition Danone Research Wageningen, The Netherlands

a r t i c l e i n f o

Article history:Received 5 March 2013Received in revised form 21 May 2013Accepted 2 July 2013Available online xxxx

Keywords:Cytokine profilesPreterm infantsNeutral and acidic oligosaccharidesOne year follow up

a b s t r a c t

Infections are common in preterm infants and cause differences in cytokine levels. Aim of this study wasto measure cytokine levels in preterm infants during the first year of life and to determine the effect offeeding a specific non-digestible carbohydrate mixture (scGOS/lcFOS/pAOS). Furthermore, other perinatalfactors in relation to these cytokine levels were analysed.

In a randomized controlled trial, preterm infants (GA <32 weeks and/or birth weight <1500 g) receiveda scGOS/lcFOS/pAOS mixture or a placebo (maltodextrin) between days 3 and 30 of life. Cytokine levels(IL-1b, IL-2, IL-4, IL-6, IL-8, IL-10, IL-17, IFN-c, and TNF-a) were analysed at 5 time points during thestudy: before start of the study, at day 7, at day 14 and at 5 and 12 months after the start of the interven-tion.

In total, 55 preterm infants in the scGOS/lcFOS/pAOS group and 58 in the placebo group were included.During the neonatal period cytokine levels increased, followed by a decrease at 5 months and 12 months.Enteral supplementation of the non-digestible oligosaccharides decreased cytokine levels at day 7 but notat day 14, indicating a temporarily anti-inflammatory effect. In the neonatal period, serious infectionbefore sampling increased all cytokine levels.

In conclusion, enteral supplementation of this specific non-digestible oligosaccharide mixturedecreased cytokine levels in preterm infants at day 7 of life, although this effect disappeared thereafter.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Preterm infants have an immature immune system whichmakes them vulnerable for infections [1]. Cytokines are a key partof the immune system, thereby playing a role in both the suscepti-bility and immune-defence of newborn infants towards infections.Cytokine levels in preterm infants differ from cytokine levels interm infants. In cord blood, pro-inflammatory cytokines such asInterleukin (IL)-2, IL-4, IL-5, IL-6, IL-8, IL-10 and tumor necrosis fac-tor a (TNF-a) are higher in preterm infants than in term infants,while there are conflicting results regarding higher or even lowerIL-1b levels in cord blood of preterm infants compared to terminfants [2,3]. Levels of pro-inflammatory cytokines such as IL-1b,IL-6, IL-8, TNF-a and IFN-c are higher in cord blood of preterminfants who were prenatally exposed to infection. IL-10, IL-18,IFN-c, TGF-b and TNF-a might differentiate between infants with

fungal and bacterial sepsis, later in the neonatal period [4]. Extre-mely low birth weight infants with blood stream infections havelower levels of IL-17 and higher levels of IL-6 and IL-8. The highestblood cytokines levels in preterm infants are found on the day ofbirth [5]. Levels of cytokines at birth or in the neonatal period inpreterm infants are possible predictors of bronchopulmonarydysplasia, white matter brain damage and cerebral palsy [6–10].

The infant’s immune system matures by exposure to intestinalmicrobiota [11,12]. This interaction leads to metabolic/immuno-logic reactions by the epithelial cells and the underlying lymphoidcells: the bacterial–epithelial ‘cross talk’ [13–15]. Preterm infantshave a delayed intestinal colonisation and possibly as a conse-quence an inappropriate bacterial–epithelial ‘crosstalk’, resultingin an inadequate maturation of the host immune defence[13,14,16,17].

In term infants, breastfeeding decreases the incidence of infec-tions [18]. Human milk oligosaccharides (HMOS) are part of the fac-tors in human milk thought to be responsible for this effect [19–21].Non-human milk oligosaccharides such as short-chain galacto-oligosaccharides (scGOS) and long-chain fructo-oligosaccharides(lcFOS) are used to mimic the functions of HMOS [22,23]. Pectin

neutral

2 J.P. van den Berg et al. / Cytokine xxx (2013) xxx–xxx

derived acidic oligosaccharides (pAOS) are able to act as receptorsanalogs and are known to inhibit the adhesion of pathogens onthe epithelial surface [24]. pAOS may also directly affect theimmune cells via interaction with selectins, dendritic cell specificC-type lectin, integrins and other target receptors such as Toll-likereceptors [23]. We recently described that, if supplemented in suf-ficient amounts, the combination of neutral oligosaccharides andacidic oligosaccharides decreases the incidence of infections in pre-term infants [25]. It was hypothesised that enteral supplementationof a scGOS/lcFOS/pAOS mixture consisting of neutral and acidic oli-gosaccharides may positively modulate the immune system.

As part of our randomized controlled trial, the aim of this studywas to measure cytokine profiles during the neonatal period andduring the first year of life of preterm infants. A unique analysis,which has never been done before in preterm infants, both duringthe neonatal period and the following year. In addition, the effectof enteral supplementation with a mixture of non-digestible neu-tral and acidic oligosaccharides but also the effect of other perina-tal factors on serum cytokine profiles were analysed.

2. Methods

2.1. Subjects

Infants with a gestational age (GA) <32 weeks and/or birthweight (BW) <1500 g, admitted to the level III NICU of the VU

Fig. 1. Trial profile. Time points T0 = <48 h after birth, T1 =

Please cite this article in press as: van den Berg JP et al. Neonatal modulation ofand acidic oligosaccharides in preterm infants. Cytokine (2013), http://dx.doi.o

University Medical Center, Amsterdam, were eligible for participa-tion in the study. Exclusion criteria were: infants with a GA > 34 -weeks, major congenital or chromosomal anomalies, death <48 hafter birth and transfer to another hospital <48 h after birth. Themedical ethical review board of the hospital approved the studyprotocol. Written informed consent was obtained from all parents.This trial was registered at isrctn.org as ISRCTN16211826.

2.2. Randomization, blinding and treatment

The infants were randomly allocated to treatment <48 h afterbirth to receive either enteral 80% scGOS/lcFOS and 20% pAOS orplacebo powder (maltodextrin) as previously described [25,26].The randomisation code was broken after data analysis was per-formed. Supplementation of the mixture or placebo was adminis-tered in increasing doses between days 3 and 30 of life to 1.5 g/kg/day to breast milk or preterm formula. Per 100 mL, the pretermformula provided 80 kcal, 2.4 g protein (casein-whey protein ratio40:60), 4.4 g fat and 7.8 g carbohydrate. The preterm formula didnot contain oligosaccharides. When infants were transferred to an-other hospital before the end of the study, the protocol was contin-ued under supervision of the principal investigator [25].

2.3. Nutritional support

Nutritional support was administered as previously described[25]. For each infant in the study a feeding schedule was proposed

day 7, T2 = day 14, T3 = 5 months and T4 = 12 months.

serum cytokine profiles by a specific mixture of anti-inflammatory neutralrg/10.1016/j.cyto.2013.07.002

J.P. van den Berg et al. / Cytokine xxx (2013) xxx–xxx 3

based on BW and the guidelines as mentioned previously [23,25].The medical staff of the NICU and the responsible paediatriciansin the regional hospitals had final responsibility for the administra-tion of parenteral nutrition and advancement of enteral nutrition.

2.4. Cytokine analysis

Blood samples were collected before the start of the interven-tion, within 48 h after birth (birth), at postnatal day 7 (day 7)and day 14 (day 14), at 4–6 weeks after the last vaccination ofthe primary series of DTaP-IPV-Hib and pneumococcal vaccina-tions at 2, 3 and 4 months (5 months) and at 4–8 weeks after theDTaP-IPV-Hib and pneumococcal booster vaccination at 11 months(12 months). Collected blood was allowed to clot at room temper-ature for 1 h and centrifuged at 3000 rpm for 10 min. Serum wascollected and stored at �80 �C until analysis. Serum samples wereanalysed for levels of IL-1b, IL-2, IL-4, IL-6, IL-8, IL-10, IL-17, IFN-cand TNF-a. Cytokine levels were measured using a fluorescentbead-based multiplex immunoassay (MIA) (Luminex xMAP tech-nology) at the National Institute for Public Health and the Environ-ment, Bilthoven, the Netherlands. Analysis was performed with aBio-Plex 200 (Bio-rad laboratories, Hercules, CA, USA). Cytokine as-say kits were purchased from Bio-rad (Hercules, CA, USA). Cytokinemeasurements were performed according to the instructions of themanufacturer with slight modifications. Samples were diluted withhuman serum diluent in a 1:1 ratio (normal 1:3). The standardcurve was calibrated on a high Photo-Multiplier Tubes (PMT)

Table 1Baseline and nutritional characteristics.a,b

Baseline characteristicsMaternal age (year)Maternal race, % Caucasian [n (%)]Obstetric diagnosis [n (%)]ChorioamnionitisPE, E or HELLPPlacental insufficiencyAntenatal antibiotics [n (%)]Antenatal corticosteroids [n (%)]Multiple birth [n (%)]Vaginal delivery [n (%)]Gestational age (weeks)Birth weight (kg)Birth weight <10th percentile [n (%)]Sex , %male [n (%)]Apgar at 5 min <6 [n (%)]pH umbilical artery <7.10 [n (%)]Surfactant medication [n (%)]PIVH [n (%)]NoneGrades I or IIGrades III or IVAntibiotics postpartum [n (%)]

Nutritional characteristicsAge at start of study supplementation (day)Time to full supplementation dose (day)Mean supplementation dose during study period (g/kg/day)Age at advancement of enteral nutrition (day)Exclusive breast milk during 30 day study period

a PE = preeclampsia; E = eclampsia; HELLP = syndrome of hemolysis, eleventricular hemorrhage.

b Student’s t test, Mann–Whitney U test and chi-square test or Fishenonparametric continuous data, respectively.

c Mean ± SD (all such valued).d Median: range in parentheses (all such values).e There were no statistically differences (p < 0.05) between both groups

Please cite this article in press as: van den Berg JP et al. Neonatal modulation ofand acidic oligosaccharides in preterm infants. Cytokine (2013), http://dx.doi.o

setting. Paired samples from a single subject were always analysedon the same plate.

2.5. Statistical analysis

The sample size of 113 infants was based on the sample size cal-culation for the primary outcome of the main trial (serious infec-tious morbidity). Normally distributed and nonparametric dataare presented as means ± SD and medians (ranges) respectively.Patient and nutritional characteristics were compared betweengroups with t test, Mann–Whitney U test, chi-square test, or Fish-er’s exact test for continuous normally distributed, nonparametriccontinuous and dichotomous data, respectively. In the primaryanalysis, generalised estimating equations (GEE) were used tocompare changes in cytokine levels over time between the groups.This method takes into account the dependency of the observa-tions within a patient and the fact that samples may not be avail-able at each measuring time. Cytokine levels below the lower limitof quantitation were assigned as half the mean lower limit of quan-titation of all plates (IL-1b 0.14, IL-2 0.06, IL-4 0.03, IL-6 0.10, IL-80.10, IL-10 0.10, IL-17 0.32, IFN-c 2.41 and TNF-a 0.23). All cyto-kine levels were expressed as geometric mean concentrations(GMC) with 95% confidence intervals (CI).

Furthermore, the effect of host and treatment related factors oncytokine levels (prenatal corticosteroids, chorioamnionitis, gender,mode of delivery, GA, serious infection <48 h before blood sample,antibiotics <48 h before blood sample, type of feeding and mildbronchopulmonary dysplasia (BPD) [23,25]) was determined by

scGOS/lcFOS/pAOS (n = 55)e Placebo (n = 58)e

32.1 ± 4.6c 31.1 ± 6.443/55 (78%) 40/58 (69%)

11/55 (20%) 13/58 (22%16/55 (31%) 18/58 (31%)4/55 (7%) 3/58 (5%)11/55 (20%) 16/58 (28%)30/55 (56%) 32/58 (56%)9/55 (16%) 13/58 (22%)31/55 (56%) 32/58 (55%)29.9 ± 1.9 29.3 ± 2.11.32 ± 0.4 1.23 ± 0.312/55 (22%) 8/58 (14%)31/55 (56%) 36/58 (62%)9/55 (16%) 5/58 (9%)2/55 (4%) 0/58 (0%)22/55 (40%) 26/58 (45%)

45/55 (82%) 46/58 (80%)8/55 (15%) 10/58 (17%)2/55 (4%) 2/58 (3%)41/55 (75%) 44/58 (76%)

2.1 (1.54–5.25)d 2.1 (1.54–3.25)11 (4–28) 11 (5–27)1.30 (0.10–1.60) 1.27 (0.17–1.79)2.8 (0.58–27.50) 2.5 (0.33–18.04)38/55 (69%) 33/58 (57%)

vated liver enzymes and low platelets; PIVH = periventricular-intra-

r’s exact test are used to analyse continuous normally distributed,

.

serum cytokine profiles by a specific mixture of anti-inflammatory neutralrg/10.1016/j.cyto.2013.07.002

Table 2Nutritional and clinical characteristics.a

scGOS/lcFOS/pAOS group (n = 55)c Placebo group (n = 58)c

Day 1 (T0) Blood samples taken at (hours) 15.35 ± 19.79b 13.54 ± 21.27n = 110 Bloodtype – Venous 3/54 (6%) 3/56 (5%)

– Capillair 4/54 (7%) 6/56 (11%)– Arterial 11/54 (20%) 17/56 (30%)– Umbilical cord 36/54 (67%) 29/56 (52%)– Missing 0/54 (0%) 1/56 (2%)

Day 7 (T1) Blood samples taken at (days) 6.98 ± 0.63 7.19 ± 0.89n = 107 Infection 648 h for sample 10/52 (19%) 13/55 (24%)

Antibiotics 648 h for sample 17/52 (33%) 24/55 (44%)Exclusive breast milk feeding 32/52 (62%) 34/55 (62%)Bloodtype – Venous 2/52 (4%) 0/55 (0%)

– Capillair 26/52 (50%) 32/55 (58%)– Arterial 21/52 (40%) 22/55 (40%)– Umbilical cord 1/52 (2%) 0/55 (0%)– Missing 2/52 (4%) 1/55 (2%)

Day 14 (T2) Blood samples taken at (days) 14.54 ± 2.03 14.44 ± 1.21n = 102 Infection 648 h for sample 1/49 (2%) 6/53 (12%)

Antibiotics 648 h for sample 21/49 (43%) 22/53 (42%)Exclusive breast milk feeding 32/49 (66) 37/53 (70%)Bloodtype – Venous 0/49 (0%) 1/53 (2%)

– Capillair 43/49 (88%) 42/53 (79%)– Arterial 4/49 (8%) 9/53 (17%)– Missing 2/49 (4%) 8/53 (15%)

5 Months (T3) Blood samples taken at (days) 176 ± 21 179 ± 21n = 88 Bloodtype – Venous 38/41 (93%) 42/47 (89%)

– Capillair 3/41 (7%) 5/47 (11%)

12 Months (T4) Blood samples taken at (days) 389 ± 23 391 ± 23n = 82 Bloodtype – Venous 31/41 (76%) 35/41 (85%)

– Capillair 9/41 (22%) 6/41 (15%)– Arterial 1/41 (2%) 0/41 (0%)

a Student’s t test, Mann–Whitney U test and chi-square test or Fisher’s exact test are used to analyse continuous normally distributed, nonparametric continuous data,respectively.

b Mean ± SD (all such valued).c There were no statistically differences (p < 0.05) between both groups.

4 J.P. van den Berg et al. / Cytokine xxx (2013) xxx–xxx

GEE analysis [27]. All statistical analysis was performed on anintention-to-treat basis. For all statistical analyses, a two-tailed pvalue <0.05 was considered significant. SPSS 20.0 (SPSS Inc., Chi-cago, IL, USA) was used for data analysis.

3. Results

Between May 2007 and November 2008, 113 of 208 eligiblepreterm infants were included in the study (Fig. 1). Baseline pa-tient and nutritional characteristics were not different in scGOS/lcFOS/pAOS (n = 55) and placebo group (n = 58) (Table 1). Nutri-tional and clinical characteristics at the time of cytokine measure-ments were not different in both groups (Table 2). In total, bloodsamples of 89% of the eligible infants at 5 months of age and 85%of the eligible infants at 12 months of age were collected (Fig. 1,and Table 2).

Blood samples were taken at 14.4 ± 20.7 h (T0), 7.1 ± 0.8 days(T1), 14.5 ± 1.6 days (T2), 177 ± 21 days (T3) and 390 ± 23 days(T4) (Table 2). In the crude GEE analysis at day 7, serum levels ofIL-1b, IFN-c and TNF-a (p = 0.003, p = 0.02, p = 0.026) were lowerafter enteral supplementation of scGOS/lcFOS/pAOS compared tothe control group. In addition, there was a trend to lower levelsfor IL-4, IL-6, IL-8 and IL-17 (p = 0.06, p = 0.09, p = 0.07, p = 0.06).At day 14, there was a trend towards lower levels of IL-1b(p = 0.06) in the scGOS/lcFOS/pAOS group compared with the pla-cebo group (Fig. 2). At 5 months, IL-1b was lower (p = 0.002) inthe scGOS/lcFOS/pAOS group and there was a trend towards lowerIL-4 (p = 0.06) in the scGOS/lcFOS/pAOS group. At 12 months IL-10

Please cite this article in press as: van den Berg JP et al. Neonatal modulation ofand acidic oligosaccharides in preterm infants. Cytokine (2013), http://dx.doi.o

was higher in the scGOS/lcFOS/pAOS group compared to the con-trol group (p = 0.03).

At day 14, there was a trend toward fewer infections in infantsin the scGOS/lcFOS/pAOS group compared with the placebo group(p = 0.06). Adjustment for serious infection and other possibleconfounders (chorioamnionitis, administration of prenatal cortico-steroids, mode of delivery, GA, gender, type of feeding, administra-tion of antibiotics, serious infection, mild BPD) did not change theresults of the primary analysis.

Data of both groups were combined for analysis of the role ofperinatal factors that may influence cytokine levels (Table 3). Seri-ous infection before sampling increased all cytokines levels (IL-1b,p = 0.02; IL-8, p = 0.01; IL-2, IL-4, IL-6, IL-10, IL-17, IFN-c and TNF-ap 6 0.001).

There was a trend toward less serious infections in the scGOS/lcFOS/pAOS group compared to the placebo group, therefore weadjusted for serious infections to analyse the change of cytokinelevels over time. In general, cytokine levels increased betweenbirth and day 7, except for IL-8. Cytokine levels at day 14 did notshow any significant difference with either day 1 or day 7, exceptfor IL-1B and IL-10 which increased from birth to day 14 (p = 0.04and p = 0.02), also after adjustment for serious neonatal infections.

Compared to the neonatal period (14 days) a decrease of IL-1b(12 months p < 0.001), IL-6 (5 months p < 0.001), IL-8 (5 monthsp < 0.001), IL-10 (5 months p = 0.002), IL-17 (5 months p = 0.02)and TNF-a (5 months p = 0.048) was seen during the follow-up.IFN-c showed an increase at 5 months (p = 0.01) but showed atrend towards a decrease at 12 months (p = 0.10). IL-2 (p = 0.07/p = 0.64) showed no significant change (Fig. 1).

serum cytokine profiles by a specific mixture of anti-inflammatory neutralrg/10.1016/j.cyto.2013.07.002

Fig. 2. Cytokine profiles. Cytokine levels in preterm infants supplemented with scGOS/lcFOS/pAOS and placebo group. Time points: before start of the study (t = 0), day 7(t = 1), day 14 (t = 2), 5 months (t = 3) and 12 months (t = 4) after birth. Generalised estimating equations (GEE) were used to assess differences between the scGOS/lcFOS/pAOS mixture and placebo group at baseline and for changes over time (compared with baseline values).

J.P. van den Berg et al. / Cytokine xxx (2013) xxx–xxx 5

Please cite this article in press as: van den Berg JP et al. Neonatal modulation of serum cytokine profiles by a specific mixture of anti-inflammatory neutraland acidic oligosaccharides in preterm infants. Cytokine (2013), http://dx.doi.org/10.1016/j.cyto.2013.07.002

Tabl

e3

Influ

ence

ofpe

rina

tal

fact

ors

oncy

toki

nele

vels

.a

Fact

orIL

-1b

IL-2

IL-4

IL-6

IL-8

IL-1

0IL

-17

IFN

-cTN

F-a

Pren

atal

cort

icos

tero

ids

1.08

(0.7

3–1.

57)a

1.32

(0.9

0–1.

95)

1.34

(0.9

6–1.

87)

1.34

(0.8

4–2.

12)

1.17

(0.8

8–1.

55)

1.37

(0.9

4–1.

98)

1.10

(0.7

8–1.

55)

1.42

(0.9

8–2.

06)

1.34

(0.8

5–2.

10)

Ch

orio

amn

ion

itis

1.29

(0.7

0–2.

13)

1.03

(0.6

1–1.

75)

0.96

(0.6

0–1.

53)

1.18

(0.6

6–2.

11)

1.14

(0.8

1–1.

60)

1.11

(0.6

4–1.

91)

0.80

(0.4

9–1.

29)

1.00

(0.5

9–1.

68)

1.31

(0.7

3–2.

36)

Gen

der

0.84

(0.5

6–1.

26)

1.20

(0.8

2–1.

77)

1.12

(0.8

0–1.

58)

1.23

(0.7

7–1.

96)

0.93

(0.7

0–1.

24)

1.37

(0.9

2–2.

04)

1.26

(0.8

7–1.

81)

1.23

(0.8

3–1.

81)

1.15

(0.7

2–1.

85)

Cae

sare

anse

ctio

n1.

00(0

.65–

1.52

)0.

95(0

.63–

1.42

)0.

85(0

.59–

1.22

)0.

81(0

.49–

1.34

)0.

79(0

.59–

1.07

)1.

01(0

.67–

1.51

)0.

84(0

.57–

1.24

)0.

85(0

.56–

1.29

)0.

84(0

.51–

1.37

)

Ges

tati

onal

age

0.98

(0.8

5–1.

13)

0.94

(0.8

3–1.

06)

0.93

(0.8

2–1.

04)

1.00

(0.8

4–1.

19)

1.03

(0.9

4–1.

14)

1.13

(0.9

9–1.

29)

0.88

(0.7

7–1.

02)

0.92

(0.8

1–1.

05)

0.90

(0.7

6–1.

07)

Infe

ctio

n6

48h

for

sam

pleb

2.16

(1.1

4–4.

09)d

4.14

(1.8

3–9.

37)c

5.56

(2.6

3–11

.77)

c6.

04(2

.54–

14.3

7)c

2.11

(1.1

7–3.

79)

d11

.41

(4.9

7–26

.20)

c7.

82(3

.25–

18.8

6)c

5.34

(2.5

9–11

.05)

c7.

95(3

.39–

18.6

7)c

An

tibi

otic

s6

48h

for

sam

ple

1.22

(0.8

4–1.

76)

1.05

(0.7

6–1.

47)

1.13

(0.8

4–1.

53)

1.57

(1.0

3–2.

42)

d1.

24(0

.92–

1.69

)1.

50(1

.07–

2.11

)d

1.26

(0.8

8–1.

80)

1.31

(0.9

3–1.

85)

1.24

(0.8

2–1.

86)

Excl

usi

vebr

east

feed

ing

0.75

(0.5

1–1.

10)

1.18

(0.8

1–1.

70)

1.20

(0.8

6–1.

67)

0.93

(0.5

9–1.

48)

1.03

(0.7

8–1.

37)

1.01

(0.6

7–1.

51)

0.89

(0.6

1–1.

29)

1.07

(0.7

3–1.

56)

0.97

(0.6

1–1.

55)

Mil

dB

PD1.

43(0

.59–

3.51

)1.

18(0

.39–

3.62

)0.

96(0

.37–

2.49

)1.

08(0

.40–

2.95

)1.

41(0

.76–

2.62

)2.

16(0

.75–

6.23

)0.

84(0

.33–

2.11

)1.

00(0

.32–

3.11

)0.

70(0

.23–

2.16

)

aD

ata

indi

cate

the

effe

ctof

afa

ctor

oncy

toki

ne

leve

ls(g

ener

alis

edes

tim

ated

equ

atio

ns)

.Th

eef

fect

can

bein

terp

rete

das

foll

ow:

inca

seof

pren

atal

cort

icos

tero

ids,

IL-1

bis

1.15

(95%

CI)

tim

esas

hig

has

wit

hou

tpr

enat

alco

rtic

oste

roid

s.C

orre

cted

for

the

infl

uen

ceof

scG

OS/

lcFO

S/pA

OS.

bSe

psis

,men

ingi

tis,

pyel

onep

hri

tis,

pneu

mon

iaor

arth

riti

sas

diag

nos

edby

aco

mbi

nat

ion

ofcl

inic

alsi

gns

and

apo

siti

vecu

ltu

re.

cp6

0.00

1.d

p<

0.05

.

6 J.P. van den Berg et al. / Cytokine xxx (2013) xxx–xxx

Please cite this article in press as: van den Berg JP et al. Neonatal modulation of serumand acidic oligosaccharides in preterm infants. Cytokine (2013), http://dx.doi.org/10.1

4. Discussion

This study shows a general increase in the cytokine levelsin the neonatal period compared to cord blood levels in pre-term infants. The precise reason for this increase is notknown but it might be a consequence of an immature im-mune system, which in turn leads to higher infectious pres-sure. After the neonatal period cytokine levels decrease asshown at 5 months and 12 months of age.

It is known that preterm infants have a more immatureimmune system compared to term infants at birth. Pretermbirth may lead to an immunological cascade resulting in anunbalanced pro-inflammatory/anti-inflammatory cytokineprofile with increased susceptibility to infections [28]. Espe-cially the first months of life may represent a critical windowin the maturation of the immune system [29]. It has beenhypothesised that each type of immune response, eitherpro- or anti-inflammatory, develops its own regulatory pop-ulation and a balanced immune response can be controlledby distinct regulatory T-cell populations.

To our knowledge this is the first study describing cyto-kine profile changes during the entire first year of life in a co-hort of preterm infants. Overall, the analyses show that afterthe neonatal period serum cytokine levels decrease until theage of one year.

Differences between preterm infants and term infantsmight be explained by differences in perinatal, host andnutrition related factors. Causes for preterm delivery suchas chorioamnionitis are known to cause changes in cytokineprofiles [30]. In our study higher levels of TNF-a were mea-sured at birth and day 7 after a pregnancy complicated bychorioamnionitis. After birth the cytokine profiles of all mea-sured cytokines in preterm infants are increased during orshortly after infections. In our study the use of antibiotics in-creased cytokine levels of IL-6 and IL-10 during the neonatalperiod. This is in line with Sood et al. who found higher cyto-kine levels during sepsis, especially in bacterial sepsis [31].

The influence of the type of feeding have been describedby Field et al. [32]. Exclusively breastfeeding decreasedcytokine levels in term infants. Kainonen et al. [33] foundan anti-inflammatory cytokine milieu in breast-fed infantscompared to formula-fed infants. Human milk oligosaccha-rides (HMOS) are thought to be responsible for at least a partof this effect [19–21]. HMOS interact with the local immunesystem of the gastro-intestinal tract [24,34,35]. The immu-nomodulatory effect of HMOS is thought to be related tothe development of a bifidogenic intestinal microbiota[11,12]. Furthermore, neutral and acidic HMOS can be ab-sorbed and cross the border membrane of the intestine, sug-gesting that HMOS may also act systemically and that theireffect is not restricted to the intestine itself. For preterm in-fants breast milk is considered to be the optimal nutrition[36–38]. In the first 10 days of life of preterm infants, exclu-sive breast milk feeding is rare, caused by delayed onset oflactation after preterm birth [36]. Previous studies show thatearly exposure to microbes induces pro-inflammatory re-sponses that shift presumed neonatal default anti-inflammatoryresponses towards a balanced pro-inflammatory/anti-inflammatory cytokine profile. In a vaccination model inmice, prebiotic oligosaccharides (scGOS/lcFOS/pAOS) shiftedthe immune response towards a more pronounced anti-inflammatory type of responsiveness. Consequences in termsof immune health and related disorders such as allergies,inflammation and infections are extensively described. Reg-ulatory cells seem to play a key role in this effect [12,23,39].

cytokine profiles by a specific mixture of anti-inflammatory neutral016/j.cyto.2013.07.002

J.P. van den Berg et al. / Cytokine xxx (2013) xxx–xxx 7

Our study shows that enteral supplementation of scGOS/lcFOS/pAOS decreased cytokine levels in preterm infants at day 7, indicat-ing a temporarily anti-inflammatory effect. This effect mimics theeffect of breast milk feeding as shown in term infants [32]. Afterday 7 this effect disappears and cytokine levels are comparable atday 14 in the scGOS/lcFOS/pAOS and the placebo group. HoweverIL-1b and IL-4 are still affected, possibly indicating an anti-allergicand anti-inflammatory stimulus. An explanation for this findingmight be that we supplemented the oligosaccharides both to breastmilk feeding and preterm formula feeding. We hypothesise thatsince preterm infants are seldom exclusively breast fed during thefirst 10 days of life, the beneficial effects of human milk oligosac-charides might be exceeded by the oligosaccharides in the scGOS/lcFOS/pAOS mixture. Another possible explanation for this mightbe that preterm infants have their peak incidence of sepsis aroundday 7. The influence of infections is larger than the influence of thescGOS/lcFOS/pAOS on the cytokine levels and therefore abolishesthis effect at day 14. This might be explained by the direct influenceof infections on cytokine levels but we hypothesise that the use ofbroad spectrum antibiotics during infections in preterm infantsdiminishes the anti-inflammatory effect by their opposite effecton the gut to neutral and acidic oligosaccharides [17,40].

Some aspects of the study design need to be addressed. An opti-mal supplementation dose of 1.5 g/kg/day was reached at a medianpostnatal age of 11 days due to restriction of the maximal osmolar-ity of the enteral feeding. Therefore, infants may not have receivedan optimal dose of scGOS/lcFOS/pAOS-supplementation to reachthe maximal effect at postnatal day 7 or 14. Furthermore, someof the cytokine levels were undetectable especially in IL-2 and IL-4, as we choose to measure unstimulated levels of cytokines to bet-ter reflect the natural cytokine response to oligosaccharide supple-mentation in preterm infants.

In summary, this study describes serum cytokine levels in pre-term infants in the first year of life. During the first week afterbirth, cytokine levels were relatively high, followed by a significantdecline later in the first year of life. External factors such as infec-tions and antibiotics influence this profile. Additionally, dietaryintervention with a unique mixture of non-digestible oligosaccha-rides (scGOS/lcFOS/pAOS) between days 3 and 30 of life decreasesmost cytokine levels at day 7, indicating a temporary anti-inflam-matory effect. This effect disappears at day 14. Most cytokines de-crease after the neonatal period (IL-1b, IL-6, IL-8, IL-10, IL-17 andTNF-a), while IFN-c increases at 5 months, indicating skewing to-wards Th1 activity, which is highly desirable at this age. The in-creased IL-10 levels at 12 months are promising and might helpin understanding the findings in other preclinical and clinical stud-ies in which the oligosaccharide mixture lead to less severe allergyand/or lower incidence of allergy related disorders [41,42].

Moreover, as major inter-individual differences in cytokine lev-els existed, future studies on the effect of these non-digestible oli-gosaccharides on the immune system should also focus on othermarkers of the immune system, such as regulatory T-cells.

Acknowledgments

We are indebted to the parents for allowing their infants to par-ticipate in the study and for participating during the first year oflife. Furthermore, we thank the medical and nursing staff of theNICU of the VU University Medical Center and all participating hos-pitals. We would like to thank P. van Gageldonk for technicalassistance.

References

[1] van der Zwet WC, Kaiser AM, van Elburg RM, Berkhof J, Fetter WP, ParlevlietGA, et al. Nosocomial infections in a Dutch neonatal intensive care unit:

Please cite this article in press as: van den Berg JP et al. Neonatal modulation ofand acidic oligosaccharides in preterm infants. Cytokine (2013), http://dx.doi.o

surveillance study with definitions for infection specifically adapted forneonates. J Hosp Infect 2005;61:300–11.

[2] Matoba N, Yu Y, Mestan K, Pearson C, Ortiz K, Porta N, et al. Differentialpatterns of 27 cord blood immune biomarkers across gestational age.Pediatrics 2009;123:1320–8.

[3] Skogstrand K, Thorsen P, Norgaard-Pedersen B, Schendel DE, Sorensen LC,Hougaard. Simultaneous measurement of 25 inflammatory markers andneurotrophins in neonatal dried blood spots by immunoassay with xMAPtechnology. Clin Chem 2005;51:1854–66.

[4] McElrath TF, Fichorova RN, Allred EN, Hecht JL, Ismail MA, Yuan H, et al. Bloodprotein profiles of infants born before 28 weeks differ by pregnancycomplication. Am J Obstet Gynecol 2011;204:24.

[5] Schelonka RL, Maheshwari A, Carlo WA, Taylor S, Hansen NI, Schendel DE, et al.T cell cytokines and the risk of blood stream infection in extremely low birthweight infants. Cytokine 2011;53:249–55.

[6] Carlo WA, McDonald SA, Tyson JE, Stoll BJ, Ehrenkranz RA, Shankaran S, et al.And neurodevelopmental outcomes in extremely low birth weight infants. JPediatr 2011;159:919–925e3.

[7] Paananen R, Husa AK, Vuolteenaho R, Herva R, Kaukola T, Hallman M. Bloodcytokines during the perinatal period in very preterm infants: relationship ofinflammatory response and bronchopulmonary dysplasia. J Pediatr2009;154:39–43e3.

[8] Rocha G, Proenca E, Guedes A, Carvalho C, Areias A, Ramos JP, et al. Cord bloodlevels IL-6, IL-8 and IL-10 may be early predictors of bronchopulmonarydysplasia in preterm newborns small for gestational age. Dis Markers2012;33:51–60.

[9] Kapitanovic Vidak H, Catela Ivkovic T, Jokic M, Spaventi R, Kapitanovic S. Theassociation between proinflammatory cytokine polymorphisms and cerebralpalsy in very preterm infants. Cytokine 2012;58:57–64.

[10] Kaukola T, Kallankari H, Tuimala J, Olsen P, Tammela O, Kingsmore SF, et al.Perinatal immunoproteins predict the risk of cerebral palsy in pretermchildren. Ann Med 2013;2011:31.

[11] Field CJ. The immunological components of human milk and their effect onimmune in infants. J Nutr 2005;135:1–4.

[12] Vos AP, M’Rabet L, Stahl B, Boehm G, Garssen J. Immune-modulatory effectsand potential working mechanisms of orally applied nondigestiblecarbohydrates. Crit Rev Immunol 2007;27:97–140.

[13] Kohler H, McCormick BA, Walker WA. Bacterial–enterocyte crosstalk: cellularmechanisms in health and disease. J Pediatr Gastroenterol Nutr2003;36:175–85.

[14] Forchielli ML, Walker WA. The role of gut-associated lymphoid tissues andmucosal defence. Br J Nutr 2005;93(Suppl 1):S41–8.

[15] Insoft RM, Sanderson IR, Walker WA. Development of immune function in theintestine and its role in neonatal diseases. Pediatr Clin North Am1996;43:551–71.

[16] Claud EC, Walker WA. Bacterial colonization, probiotics, and necrotizingenterocolitis. J Clin Gastroenterol 2008;42(Suppl 2):S46–52.

[17] Westerbeek EA, van den Berg A, Lafeber HN, Knol J, Fetter WP, van Elburg RM.The bacterial colonisation in preterm infants: a review of the literature. ClinNutr (Edinburgh, Scotland) 2006;25:361–8.

[18] Duijts L, Ramadhani MK, Moll HA. Breastfeeding protects against infectiousdiseases during infancy in industrialized countries. A systematic review.Matern Child Nutr 2009;5:199–210.

[19] Oddy WH. The impact of breastmilk on infant and child health. Breastfeed Rev2002;10:5–18.

[20] McGuire W, Anthony MY. Donor human milk versus formula for preventingenterocolitis in preterm infants: systematic review. Arch Dis Child FetalNeonatal Ed 2003;88:F11–4.

[21] Holt PG, Sly PD, Bjorksten B. Atopic versus infectious diseases in childhood: aof balance? Pediatr Allergy Immunol 1997;8:53–8.

[22] Gnoth MJ, Rudloff S, Kunz C, Kinne RK. Investigations of the in vitro transportof human milk oligosaccharides by a Caco-2 monolayer using a novel highperformance liquid chromatography-mass spectrometry technique. J BiolChem 2001;276:34363–70.

[23] Boehm G, Moro G. Structural and functional aspects of prebiotics used ininfant. J Nutr 2008;138:1818S–28S.

[24] Boehm G,SB. Oligosaccharides. In: Mattila-Sandholm Te, editor. Functionaldairy products. Cambridge: Woodhead; 2002. p. 203–43.

[25] Westerbeek EA, van den Berg JP, Lafeber HN, Fetter WP, Boehm G, Twisk JW,et al. Neutral and acidic oligosaccharides in preterm infants: a randomized,double-blind, placebo-controlled trial. Am J Clin Nutr 2009;91:679–86.

[26] Westerbeek EA, van Elburg RM, van den Berg A, van den Berg J, Twisk JW,Fetter WP, et al. Design of a randomised controlled trial on immune effects ofacidic and neutral oligosaccharides in the nutrition of preterm infants: carrotstudy. BMC Pediatr 2008;8:46.

[27] Jobe AH, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med2001;163:1723–9.

[28] Sadeghi K, Berger A, Langgartner M, Prusa AR, Hayde M, Herkner K, et al.Immaturity of infection control in preterm and term newborns isassociated with impaired toll-like receptor signaling. J Infect Dis2007;195:296–302.

[29] Singhal A. Early nutrition and long-term cardiovascular health. Nutr Rev2006;64. S44-; discussion S72-91.

[30] McElrath TF, Fichorova RN, Allred EN, Hecht JL, Ismail MA, Yuan H, et al. Bloodprotein profiles of infants born before 28 weeks differ by pregnancycomplication. Am J Obstet Gynecol 2011;204:418e1–418e12.

serum cytokine profiles by a specific mixture of anti-inflammatory neutralrg/10.1016/j.cyto.2013.07.002

8 J.P. van den Berg et al. / Cytokine xxx (2013) xxx–xxx

[31] Sood BG, Shankaran S, Schelonka RL, Saha S, Benjamin Jr DK, Sanchez PJ, et al.Cytokine profiles of preterm neonates with fungal and bacterial sepsis. PediatrRes 2012;72:212–20.

[32] Field CJ, Van Aerde JE, Robinson LE, Clandinin MT. Effect of providing a formulasupplemented with long-chain polyunsaturated fatty acids on immunity infull-term neonates. Br J Nutr 2008;99:91–9.

[33] Kainonen E, Rautava S, Isolauri E. Immunological programming by breast milkcreates an anti-inflammatory cytokine milieu in breast-fed infants comparedto formula-fed infants. Br J Nutr 2012:1–9.

[34] Arslanoglu S, Moro GE, Boehm G. Early supplementation of prebioticoligosaccharides protects formula-fed infants against infections during thefirst 6 months of life. J Nutr 2007;137:2420–4.

[35] Bode L. Recent advances on structure, metabolism, and function of humanmilk oligosaccharides. J Nutr 2006;136:2127–30.

[36] Corpeleijn WE, Kouwenhoven SM, Paap MC, van Vliet I, Scheerder I, Muizer Y,et al. Intake of own mother’s milk during the first days of life is associated withdecreased morbidity and mortality in very low birth weight infants during thefirst 60 days of life. Neonatology 2012;102:276–81.

Please cite this article in press as: van den Berg JP et al. Neonatal modulation ofand acidic oligosaccharides in preterm infants. Cytokine (2013), http://dx.doi.o

[37] Hylander MA, Strobino DM, Dhanireddy R. Human milk feedings and infectionamong very low birth weight infants. Pediatrics 1998;102:E38.

[38] Lucas A, Cole TJ. Breast milk and neonatal necrotising enterocolitis. Lancet1990;336:23.

[39] van’t Land B, Schijf M, van Esch BC, van Bergenhenegouwen J, Bastiaans J,Schouten B, et al. Regulatory T-cells have a prominent role in the immunemodulated vaccine response by specific oligosaccharides. Vaccine2010;28:5711–7.

[40] Bennet R, Eriksson M, Nord CE, Zetterstrom R. Fecal bacterial microflora ofnewborn infants during intensive care management and treatment with fiveantibiotic regimens. Pediatr Infect Dis 1986;5:533–9.

[41] Gruber C, van Stuijvenberg M, Mosca F, Moro G, Chirico G, Braegger CP, et al.Occurrence of early atopic dermatitis because of immunoactive prebioticsamong low-atopy-risk infants. J Allergy Clin Immunol 2010;126:791–7.

[42] Schouten B, van Esch BC, Hofman GA, Boon L, Knippels LM, Willemsen LE, et al.Oligosaccharide-induced whey-specific CD25(+) regulatory T-cells areinvolved in the suppression of cow milk allergy in mice. J Nutr2010;140:835–41.

serum cytokine profiles by a specific mixture of anti-inflammatory neutralrg/10.1016/j.cyto.2013.07.002