Nutrition Journal of Parenteral and Enteral

0 downloads 0 Views 523KB Size Report
Jan 1, 2011 - 12); (4) PN plus colostrum MEN followed by colostrum feeding ... umbilical catheter with PN solution (Nutriflex Liquid plus ... g/kg; fat, 51 g/kg).
Journal of Parenteral and Enteral Nutrition http://pen.sagepub.com/

Diet-Dependent Effects of Minimal Enteral Nutrition on Intestinal Function and Necrotizing Enterocolitis in Preterm Pigs Malene Skovsted Cilieborg, Mette Boye, Thomas Thymann, Bent Borg Jensen and Per Torp Sangild JPEN J Parenter Enteral Nutr 2011 35: 32 DOI: 10.1177/0148607110377206 The online version of this article can be found at: http://pen.sagepub.com/content/35/1/32

Published by: http://www.sagepublications.com

On behalf of:

The American Society for Parenteral & Enteral Nutrition

Additional services and information for Journal of Parenteral and Enteral Nutrition can be found at: Email Alerts: http://pen.sagepub.com/cgi/alerts Subscriptions: http://pen.sagepub.com/subscriptions Reprints: http://www.sagepub.com/journalsReprints.nav Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Jan 11, 2011 What is This?

Downloaded from pen.sagepub.com by guest on October 11, 2013

Original Communication

Diet-Dependent Effects of Minimal Enteral Nutrition on Intestinal Function and Necrotizing Enterocolitis in Preterm Pigs

Journal of Parenteral and Enteral Nutrition Volume 35 Number 1 January 2011 32-42 © 2011 American Society for Parenteral and Enteral Nutrition 10.1177/0148607110377206 http://jpen.sagepub.com hosted at http://online.sagepub.com

Malene Skovsted Cilieborg, Msc1,2; Mette Boye, PhD2; Thomas Thymann, PhD1; Bent Borg Jensen, PhD3; and Per Torp Sangild, PhD, DVSc, DMSc1 Financial disclosure: The study was supported by funding by the Danish Research Councils under the program of FØSU.

Background: A rapid advance in enteral feeding is associated with necrotizing enterocolitis (NEC) in preterm infants. Therefore, minimal enteral nutrition (MEN) combined with parenteral nutrition (PN) is common clinical practice, but the effects on NEC and intestinal function remain poorly characterized. It was hypothesized that a commonly used MEN feeding volume (16–24 mL/kg/d) prevents NEC and improves intestinal structure, function, and microbiology in preterm pigs. Methods: After preterm birth pigs were stratified into 4 nutrition intervention groups that received the following treatments: (1) PN followed by full enteral formula feeding (OF group, n = 12); (2) PN supplemented with formula MEN and followed by full formula feeding (FF, n = 12); (3) PN plus colostrum MEN followed by formula feeding (CF, n = 12); (4) PN plus colostrum MEN followed by colostrum feeding (CC, n = 10). Results: NEC was absent in the CC group but frequent in the other groups (50%–67%). Compared with other

groups, CC pigs showed improved mucosal structures, brush border enzyme activities, and hexose absorption (all P < .05). Relative to formula MEN, colostrum MEN thus improved gut function but did not prevent later formula-induced gut dysfunction and NEC. However, in CF pigs, intestinal lesions were restricted to the colon, compared with all regions in OF and FF pigs, which indicated proximal protection of colostrum MEN. Bacterial composition was not affected by MEN, diet, or NEC outcomes, but bacterial load and concentrations of short-chain fatty acids were reduced in the MEN groups. Conclusion: Colostrum MEN improves intestinal structure, function, and NEC resistance in preterm pigs but does not protect against gut dysfunction and NEC associated with later full enteral formula feeding. (JPEN J Parenter Enteral Nutr. 2011;35:32-42)

Clinical Relevancy Statement

volume and nature of MEN diet are lacking. Using a clinically relevant preterm pig model, we show that MEN doses similar to those often used for infants were safe and did not increase the risk of NEC. On the other hand, MEN did not protect against NEC when followed by full enteral feeding with formula. Only when bovine colostrum was used for MEN and as the full enteral feed was intestinal maturation stimulated and NEC absent. Our study support the clinical practice to initiate enteral feeding immediately after birth with a NEC-protective diet like mother’s milk or colostrum. We need more studies to clarify the optimal diet and feeding advancement after preterm birth, especially when mother´s milk is not available.

Keywords:   enteral feeding; preterm neonates; animal model

In the feeding of preterm infants, the goal is to achieve adequate body growth and organ development without increasing the risk of metabolic complications and NEC. MEN feeding is believed to help the transition from TPN to full enteral feeding but data on the optimal timing, From the 1Department of Human Nutrition, Faculty of Life Science, University of Copenhagen, Denmark; 2National Veterinary Institute, Technical University of Denmark, Copenhagen, Denmark; and 3Department of Animal Health and Bioscience, Faculty of Agricultural Sciences, University of Aarhus, Tjele, Denmark. Received for publication October 23, 2009; accepted for publication April 6, 2010.

Introduction

Address correspondence to: Per Torp Sangild, PhD, DVSc, DMSc, Department of Human Nutrition, Faculty of Life Science, University of Copenhagen, 30 Rolighedsvej, DK-1958 Frederiksberg C, Denmark; e-mail [email protected].

The transition from parenteral nutrition (PN) to enteral nutrition (EN) at birth is particularly challenging for preterm infants. The gastrointestinal (GI) tract is structurally 32

Minimal Enteral Nutrition and Necrotizing Enterocolitis / Cilieborg et al   33

and functionally immature at this time, with decreased gut motility and low digestive and absorptive capacity. Therefore, the preterm infant is at risk of overfeeding, with resulting food intolerance and bacterial overgrowth.1,2 Together with immature and uncontrolled immune responses, this may predispose to necrotizing enterocolitis (NEC), a serious condition of intestinal inflammation and tissue necrosis associated with increased risk of morbidity and mortality. Enteral feeding precedes NEC, and a common preventive strategy has been to nourish preterm infants via PN after birth. Luminal food stimulation, as occurs in late pregnancy when the fetus swallows amniotic fluid, may be required for proper structural and functional maturation of the gut for preterm newborns.1,3 Delayed enteral feeding following 2 to 3 days of PN thus increases enteral feeding–induced NEC in preterm pigs.4 Other studies have documented that PN may be associated with impaired intestinal growth and mucosal atrophy,5-7 impaired intestinal permeability,8 decreased digestive and absorptive capacity,9 more pathogenic microbiota,10 and systemic inflammation.11-13 Against this background, a combination of PN and minimal EN (MEN) has been advocated for newborn preterm infants, but protocols vary greatly regarding the timing of enteral feeding, dose, volume advancement, and type of diet (ie, formula or mother’s milk), and the effects on NEC prevention remain uncertain, despite extensive research.14-16 It is also unclear whether MEN affects intestinal colonization, whereas both diet17-19 and prematurity20,21 are believed to affect the gut microbiota and following NEC development.20-25 Through a variety of bioactive components that stimulate and modulate growth, immunity, and intestinal colonization,13,18,19 breast milk and colostrum are more likely to improve the effects of MEN, when compared with formula. We hypothesized that MEN feeding improves NEC resistance in preterm neonates via beneficial effects on intestinal structure, function, and microbiology, and we further predicted that MEN with colostrum would be superior to MEN with formula. To test our hypotheses, we used a previously described animal model with preterm pigs in which diet-dependent development of NEC occurs in 25% to 60% of animals.4 Clinically relevant volumes of MEN were applied during an initial 2-day PN period and after a subsequent 2-day period with full enteral feeding; our end points included gut structure, digestive and absorptive function, and a detailed microbiological characterization.

Methods Animal Protocol and Experimental Design Forty-six pigs from 3 sows (Large White × Danish Landrace) were delivered via caesarean section at 90% gestation, immediately transferred to our perinatal facilities,

and reared in temperature-regulated and oxygen-regulated incubators (Air-Shields; Hatboro, PA). Pigs were fitted with orogastric (6 Fr; Pharmaplast, Roskilde, Denmark) and umbilical catheters (4 Fr; Portex, Kent, UK) and received 12 mL/kg of their mother’s serum intraarterially during the first 24 hours to obtain immunologic protection. All pigs were initially fed intraarterially through the umbilical catheter with PN solution (Nutriflex Liquid plus; B. Braun, Melsungen, Germany; 4 mL/kg/h advancing to 6 mL/kg/h) with added amino acids (Vamin; Fresenius Kabi, Uppsala, Sweden) to meet requirements of pigs: energy, 3,123 kJ/L (300–450 kJ per kg of body weight per day); amino acids, 45 g/L (4.3–6.5 g/kg/d); glucose, 72 g/L (6.9–10.4 g/kg/d); lipids (equal amounts of soybean oil and medium-chain triglycerides [MCTs]), 31 g/L (3.0–4.5 g/kg/d), osmolality; 1,540 mosmol). During the initial 2-day period of PN, the pigs received either saline or MEN corresponding to 18% of their daily energy intake (day 1, 2 mL/kg every 3 hours; day 2, 3 mL/ kg every 3 hours) as either formula (mixed from Pepdite, Maxipro, and Liquigen-MCT, kindly donated by Nutricia, Allerød, Denmark; energy, 4,140 kJ/kg; protein, 64 g/kg; carbohydrate, 45 g/kg; fat, 61 g/kg) or cow’s colostrum (energy, 5,297 kJ/kg; protein, 176 g/kg; carbohydrate, 22 g/kg; fat, 51 g/kg). Colostrum was diluted in tap water 2:1 to adjust it to the energy content of formula. After 2 days on PN, pigs were fed full EN with formula or colostrum (15 mL/kg every 3 hours, corresponding to ~50% of the energy requirement of a full-term newborn pig) for another 2 days before euthanasia (via intravenous [IV] pentobarbitone 60 mg/kg) and tissue collection. The above procedures constitute the standard protocol for our preterm pig model previously described in detail4,26 and was approved by the National Committee on Animal Experimentation in Denmark. The pigs were stratified according to birth weight into 4 nutrition intervention groups. One group received formula MEN boluses followed by full enteral formula feeding (FF, n = 12), a second group received colostrum MEN followed by full enteral formula feeding (CF, n = 12), a third group received colostrum in both periods (CC, n = 10), and a control group received oral saline corresponding to the MEN boluses and were fed formula during the full enteral period (OF, n = 12). As previously described,27 an in vivo sugar absorption test was performed before and 30 hours after the transition to full enteral feeding. A baseline blood sample was collected before administration of the oral test bolus (15 mL/kg, 5% galactose and 2% mannitol) and again at 20, 40, and 60 minutes after the bolus.

Clinical Evaluation and Tissue Collection After introduction of enteral feeding, the pigs were kept under surveillance for appearance of NEC symptoms

34   Journal of Parenteral and Enteral Nutrition / Vol. 35, No. 1, January 2011

such as bloody diarrhea, lethargy, respiratory distress, and abdominal distension. If NEC symptoms appeared, the pigs were immediately euthanized, and tissue was collected, whereas the remaining pigs were euthanized 25 to 36 hours after introduction of full enteral feeding. Five GI regions (stomach; proximal, middle, and distal small intestine; and colon) were scored for NEC lesions, according to a macroscopic NEC evaluation system, where 1 = normal, 2 = local hyperemia and edema, 3 = hyperemia, extensive edema, and local hemorrhage, 4 = extensive hemorrhage, 5 = local necrosis and pneumatosis intestinalis, and 6 = extensive necrosis and pneumatosis intestinalis. NEC was defined as a score of 3 or greater in at least 1 region.4 Organs were weighed, and tissue samples from 3 regions in the small intestine together with stomach and colon contents were collected and snap frozen for later functional and microbiological analyses. Tissue samples from the distal small intestine were formaldehyde fixed and paraffin embedded for later histological and fluorescence in situ hybridization (FISH) analyses.

Mucosal Structure and Digestive Function Mucosa weight proportions were calculated after scraping off the mucosa from 10-cm sections of the small intestine regions, and the proportion of mucosa was determined based on both wet and dry weights. Mucosal morphology was further characterized by measurements of villus height and crypt depth from pictures obtained from paraffin-embedded sections of distal small intestine scanned using an arrayWoRxe microarray scanner (Applied Precision, Issaquah, WA) and the morphometric software softWoRx Explorer 1.1 (Applied Precision). Assays for mucosal activity of disaccharidases (maltase, sucrase, and lactase) and peptidases (aminopeptidase N[ApN], aminopeptidase A[ApA], and dipeptidyl peptidase IV[DPP-IV]) were performed on small intestine tissue homogenates (homogenized in 1% Triton X-100) using specific substrates, as described previously.26 In vivo sugar absorption was determined by measuring plasma levels of D-galactose and mannitol using kits based on galactose dehydrogenase and mannitol dehydrogenase, respectively.27

Microbiology From the time of birth, gastric aspirates (1–2 mL) were collected daily via the orogastric catheter for microbiological analyses. The first sample was taken on day 1 before any MEN feeding, whereas samples were collected before a MEN bolus on day 2. The third sample was taken 2.5 hours after the first bolus of full enteral feeding on day 3, and sample 4 was obtained at tissue collection. The

aspirates were plated for enumeration of colony-forming units (CFUs) on unselective calf blood agar (SSI Diagnostika, Hillerød, Denmark), Clostridium perfringens–selective tryptose–sulfite–cycloserine (TSC) agar (TSC plates; National Veterinary Institute, Copenhagen, Denmark) and lactic acid bacteria–selective de Man, Rogosa, and Sharpe28 (MRS) agar (MR; National Veterinary Institute, Copenhagen, Denmark) and incubated anaerobically. Finally, pH was measured on undiluted aspirate samples using pH strips with pH interval 0–6 (Merck Chemicals, Damstadt, Germany). The small intestine microbiota was characterized by terminal restriction fragment length polymorphism (T-RFLP) as described previously.29,30 In short, DNA extractions were made on frozen tissue sections of the distal small intestine using the QIAamp DNA Mini Kit (Qiagen, Ballerup, Denmark) and polymerase chain reaction (PCR) amplified using 16S rDNA primers S-D-Bact0008-aaa-S-20; 5′-AGAGTTTGATCMTGGCTCAG-3′ labeled with 5′FAM and S-D-Bact-0926-a-A-20; 5′-CCGTCAATTYMTTTRAGTTT-3′ (DNA Technology, Aarhus, Denmark).30 PCR products were digested with restriction enzyme CfoI and sequenced (Applied Biosystems Genetic Analyzer 3130/3130xl, Nærum, Denmark). Data analysis was performed with BioNumerics 4.5 software (Applied Maths, Kortrijk, Belgium) aligning terminal restriction fragments (T-RFs) against an internal standard. For standardization, relative intensity of individual T-RFs was calculated as the T-RF intensity divided by the total sample intensity. Only dominating T-RFs with a mean relative intensity >0.01 were chosen for graphic presentation. Identification of the dominating T-RFs were proposed based on in silico digest performed with Microbial Community Analysis III (MICA) online software (http://mica.ibest.uidaho.edu/digest.php; University of Idaho, Moscow, ID) and the Ribosomal Database Project II database (release 9, update 37, Bacterial SSU 16S recombinant RNA; Michigan State University, East Lansing, MI). Bacterial microcolonies were visualized by FISH on 3-µm paraffin-embedded cross-sections of distal small intestine sections using a CY3-labeled general bacteria 16s rDNA probe (5′-GCTGCCTCCCGTAGGAGT-3′31 and an arrayWoRxe scanner. According to bacterial abundance and localization, each tissue section was given a score on a 4-point scale, where 1 = no visible bacteria, 2 = few microcolonies, 3 = abundant bacteria located in the mucosal periphery or in the intestinal lumen, and 4 = extensive colonization and bacteria located deep between the villi or even translocating the epithelial lining. Bacterial metabolic activity was investigated by measuring the concentrations of 16 different short-chain fatty acids (SCFAs) in the stomach and colon contents using gas chromatography, as previously described.31

Minimal Enteral Nutrition and Necrotizing Enterocolitis / Cilieborg et al   35

OF

5

NEC scores

4 3

FF

CF

CC

a a

a

a

a

a

a,b a,b

a,b b

2 b,c

1 0

b

Stomach

c

b

c

b

Small intestine

Colon

Average

Figure 1.  Necrotizing enterocolitis (NEC) scores (mean ± standard error of the mean) in gastrointestinal organs, based on macroscopic tissue evaluation. Means not sharing the same superscript letter within regions are significantly different (P < .05). CC, group receiving colostrum in both periods; CF, group receiving colostrum minimal enteral nutrition followed by full enteral formula feeding; FF, group receiving formula in both periods; OF, control group receiving oral saline during the PN period and formula during the full enteral feeding period.

Table 1.   Relative Weight and Mucosal Structures of the Small Intestine of Preterm Pigs Treatment Groups OF Relative SI weight, g/kg Villus height, µm Villus/crypt ratio SI circumference, mm SI dry-weight proportion, % Mucosa proportion, %

FF a,b

29.3 ± 1.5 400.6 ± 61.7a 4.1 ± 0.5a 13.0 ± 0.7a 14.9 ± 0.9a,b 65.9 ± 2.1

CF a,b

30.6 ± 2.5 482.5 ± 79.9a,b 4.7 ± 0.8a 12.2 ± 0.7a 13.9 ± 0.7b 65.0 ± 3.4

CC a

30.8 ± 1.5 452.8 ± 47.7a,b 4.6 ± 0.5a 11.7 ± 0.4a,b 16.0 ± 0.3a 67.4 ± 1.8

24.7 ± 0.8b 591.1 ± 31.8b 6.6 ± 0.4b 10.4 ± 0.4b 15.3 ± 0.5a 66.6 ± 1.8

CC, group receiving colostrum in both periods; CF, group receiving colostrum minimal enteral nutrition followed by full enteral formula feeding; FF, group receiving formula in both periods; OF, control group receiving oral saline during the PN period and formula during the full enteral feeding period; SI, small intestine. Values are mean ± standard error of the mean. Mucosa proportion was calculated as dry weight mucosa divided by dry weight total SI. Means not sharing the same superscript letter are significantly different (P < .05).

Calculations and Statistical Analyses All results are presented as mean ± standard error of the mean. Group differences in NEC incidences were tested using Fisher exact test (SAS/STAT 9.1; SAS Institute, Cary, NC). Group differences in mean NEC score, FISH score, villus height, crypt depth, and villus/crypt ratio were tested with 2-way analysis of variance using the PROC MIXED procedure of SAS, with treatment, diet, and sex as fixed variables and pigs and sows as random variables. Differences in plasma galactose and mannitol levels were tested with an incorporated repeated-measures approach in the PROC MIXED procedure. Group differences in bacterial enumeration were tested with Student t test. Differences in mean relative T-RF intensity were tested using Monte Carlo estimates for the Wilcoxon 2-sample exact test in SAS. Based on overall similarity of all T-RFs, principal component analysis

(PCA) was generated in BioNumerics 4.5 (Applied Maths, Kortrijk, Belgium). For all analyses, P < .05 was used as the critical level of significance.

Results Clinical Outcomes There were no differences among treatment groups in birth weight (1,058 ± 39 g), weight gain (27 ± 2 g), age (73.1 ± 2.7 hours), or rectal temperatures (0 hours, 34.8 ± 0.1; 6 hours, 37.8 ± 0.2; and 24 hours, 38.5°C ± 0.3°C). Eleven hours after initiation of full enteral feeding, NEC was observed in 2 OF pigs and 1 FF pig, and they were euthanized for tissue collection. The remaining pigs were euthanized for tissue collection between 25 and 36 hours after full EN. NEC was absent in CC pigs (0/10),

36   Journal of Parenteral and Enteral Nutrition / Vol. 35, No. 1, January 2011

Sucrase

0.6 0.5 0.4

0.3

a,c

a,c

a,b

a,b a

a,b

8.0 a,b

a

6.0 2.0

0.0

0.0

c

a

5.0 b,c

4.0 b

b

b a,b

a a

3.0 2.0

a a

b

c a,b

b

b a

a

a

a

1.0

a,b a,b

a a

a,b

a

a

a

b a,b

2.0

0.5

0.5

OF

b b

1.5 1.0

Mid SI

c

2.5

1.0

Proximal SI

DPP-IV

3.0

a

2.5

0.0

b b

a

Maltase

1.5

a

a

0.0

3.0

2.0

a

a

6.0

c

a

a,b b

b

ApA

Lactase

b

b

b

4.0

0.1

80 70 60 50 40 30 20 10 0

b c

10.0

b

c

0.2

U/g

a

ApN

12.0

b

Distal SI

FF

0.0

a a

a,c

a

b a

a

a,b

c

Proximal SI

CF

Mid SI

Distal SI

CC

Figure 2.   Activity of brush border enzymes (sucrase, lactase, maltase, aminopeptidases A and N [ApA and ApN], and intravenous dipeptidyl peptidase [DPP-IV]) in proximal, middle, and distal small intestine (SI) (mean ± standard error of the mean). Means not sharing the same superscript letter within region differ significantly (P < .05). CC, group receiving colostrum in both periods; CF, group receiving colostrum minimal enteral nutrition followed by full enteral formula feeding; FF, group receiving formula in both periods; OF, control group receiving oral saline during the PN period and formula during the full enteral feeding period.

which was significantly different from other groups, in which incidences were 67% (8/12) for OF pigs, 58% (7/12) for FF pigs, and 50% (6/12) for CF pigs. There were severe NEC lesions throughout all 5 GI regions in the OF and FF groups, whereas only the colon was affected in the CF group (Figure 1). Relative to body weight, in the CC group, the weight of the colon was significantly lower, and the weight of the stomach, small intestine, and the length of the small intestine tended to be lower, compared with the mean of the values from the three other groups, which were not different (colon, 6.3 ± 0.9 vs 9.0 ± 0.7 g/kg, P = .05; stomach, 4.9 ± 0.2 vs 6.4 ± 0.4 g/kg, P = .16; small intestine, 24.7 ± 0.8 vs

29.4 ± 1.4 g/kg, P = .12; small intestine length, 235.0 ± 31.4 vs 276.7 ± 14.2 cm/kg, P = .5). Relative weights of other organs did not differ among the treatment groups (overall means: lungs, 20.0 ± 0.7; spleen, 1.8 ± 0.1; heart, 7.5 ± 0.2; kidneys, 9.0 ± 0.2, brain, 21.1 ± 1.9) except that CC pigs had reduced liver weights compared with the other groups (25.4 ± 1.2 vs 28.4 g/kg, P < .05).

Mucosal Structure and Digestive Function A moderate treatment effect was seen in dry/wet weight proportions of the small intestine, with the lowest values in FF pigs, indicating a greater degree of tissue edema

Minimal Enteral Nutrition and Necrotizing Enterocolitis / Cilieborg et al   37

A

B

1,500 1,250

µmol/L

1,000 750 500 250

C

0

D

300 250

mg/L

200 150 100 50 0 Baseline

20 min OF

FF

40 min

60 min

CF+CC

Baseline OF

20 min FF

40 min CF

60 min CC

Figure 3.   Increments of plasma galactose and mannitol at 20, 40, and 60 minutes after administration of an oral test bolus (mean ± standard error of the mean). (A) Plasma galactose before enteral feeding. (B) Plasma galactose 24 hours after transition to enteral feeding. (C) Plasma mannitol before enteral feeding. (D) Plasma mannitol 24 hours after transition to enteral feeding. Before full EN (A and C), CF and CC pigs were pooled because both groups were fed colostrum MEN. Before enteral feeding (A and C), repeated measurements statistical analyses revealed significantly lower galactose and mannitol levels in FF pigs compared with other groups, whereas CC pigs had significantly higher galactose and mannitol levels 24 hours after enteral feeding compared with other groups (B and D), P < .05. CC, group receiving colostrum in both periods; CF, group receiving colostrum minimal enteral nutrition followed by full enteral formula feeding; FF, group receiving formula in both periods; OF, control group receiving oral saline during the PN period and formula during the full enteral feeding period.

(Figure 2). Circumference of the small intestine, which inversely indicates submucosal muscle tone, was significantly smaller in the proximal and distal segments in CC pigs than in OF and FF pigs (Table 1). Circumference was also significantly decreased in healthy pigs compared with NEC pigs (average of proximal and distal segments, 10.8 ± 0.3 vs 12.8 ± 0.4 mm, P < .001). Improved gut structure for CC pigs was supported by histomorphological measurements on tissue sections of the distal small intestine, with higher villi than in other groups (591 ± 32 vs 446 ± 35μm, P = .05) and higher villus/crypt ratios (Table 1). CC pigs generally had increased digestive enzyme activities in the small intestine and particularly in the distal segment (Figure 2). Among disaccharidases, lactase activity represented the most pronounced group difference and was significantly higher in CC pigs across all 3 small intestine regions. Activity of ApA, ApN, and DPP-IV was significantly higher in CC pigs in the distal region than in other diet groups (Figure 2). FF and CF pigs had intermediate activities of digestive enzymes, whereas the

lowest activity levels were generally found in OF pigs, though the level of significance varied (Figure 2). Plasma levels of galactose and mannitol after an oral test bolus revealed that pigs fed formula during MEN (FF pigs) had decreased in vivo absorption of the 2 sugars compared with other groups (Figure 3). Thirty hours after transition to full EN, the in vivo absorption of both galactose and mannitol decreased in all groups, except the CC group (Figure 3).

Microbiology Gastric aspirates were obtained from 20 to 28 pigs during the first 3 days. Before the first oral feeding on day 1, 9 of 20 gastric samples were positive for bacteria, with numbers from 20 to 64, 000 CFU/mL (mean 2.1 × 103) across all cultivation media. After 4 MEN feedings on day 2, all samples except from 1 OF pig were positive for bacteria, with no significant differences in bacterial numbers among diet groups or incubation media (mean 6.9 × 106;

38   Journal of Parenteral and Enteral Nutrition / Vol. 35, No. 1, January 2011 Stomach pH 5

pH

4.5 4 3.5

Stomach microbiology

CFU/mL (log10)

10

8

6

4

2 Day 1

Day 2

Day 3

Day 4

Blood aerobic

TSC anaerobic

Blood anaerobic

MRS anaerobic

Figure 4.   Mean pH and bacterial numbers of stomach contents (pooled means across the 4 treatment groups). Bacteria were cultivated on unselective calf blood agar, de Man, Rogosa, and Sharpe28 (MRS), and tryptose–sulfite–cycloserine (TSC) media selecting for Lactobacilli and Clostridium perfringens, respectively.

Figure 4). Numbers further increased on days 3 and 4 to an average of 5.6 × 107 and 8.06 × 107 CFU/mL, respectively, with no significant differences among treatment groups (Figure 4). Clinical status (NEC or healthy) was not associated with bacterial counts on day 4 (data not shown). There was no difference in the pH of stomach contents among treatment groups, except on day 4, when pH in CF pigs was significantly lower than in OF and CC pigs (3.5 ± 0.2 vs 4.4 ± 0.2 and 4.2 ± 0.3, respectively). Average pH across treatment groups remained constant during the experimental period, with a transient increment after transition to full EN on day 3 (Figure 4). Based on the distribution and intensity of T-RFs, PCA analysis presented spatial separation of pigs according to overall microbiota similarities but did not reveal any differences in distal intestine tissue bacterial profile among diet groups or between NEC and healthy pigs (data not shown). Only 4 of the dominating T-RFs were significantly different

among groups (Figure 5): OF pigs displayed the lowest and CF pigs the highest intensity of T-RF 61, possibly representing Lactococcus lactis, as indicated by the in silico MICA digest. CC pigs displayed the highest intensity of T-RF 254, possibly Lactobacillus delbrueckii and T-RF 500 with unknown identity, and the lowest intensity of T-RF 584, possibly a Lactococcus. Colonization differences were not seen between healthy pigs and pigs with NEC, except for T-RF 584, with a significantly higher abundance in pigs with NEC. Contradicting previous studies,4,26,30 abundance of C perfringens, represented by T-RF 233, was not altered by diet or disease status (Figure 5). Average FISH scores expressing the in situ abundance and location of bacteria in the distal small intestine were significantly higher in OF pigs than in other groups (Figure 6) and in pigs diagnosed for NEC and healthy pigs, respectively (2.7 ± 0.2 vs 1.8 ± 0.2, P < .01). Among the 16 different SCFAs investigated, only acetate, butyrate, lactate, succinate, caproate, and octanoate were detected in stomach contents, where lactate and octanoate accounted for most of the total SCFA measured (Figure 7). A slightly more diverse microbiota in the colon was indicated by the presence of the same and 6 additional SCFAs (formate, propionate, iso-butyrate, isovalerate, valerate, and isocaproate), where lactate, acetate, and succinate dominated (Figure 7). Total concentrations of SCFAs reflecting the overall bacterial metabolic activity and thus bacterial load were highest for OF pigs in both regions and decreased gradually for FF and CF pigs, to reach the lowest values in CC pigs (Figure 7).

Discussion The use of MEN is common in clinical care of preterm infants, but its effects on GI maturation and NEC sensitivity remain poorly defined.15,33,34 Use of a preterm pig model enabled us to clarify the effects of MEN on spontaneous NEC development and some key parameters of GI structure, function, and microbiology. Our results show that the effects of MEN, at doses similar to those applied in the human neonatal clinics (16–24 mL/kg/d, corresponding to 18% of daily energy intake), are highly diet-dependent, with improved effects of colostrum vs formula. A high incidence of NEC (50%–67%) remained present in all groups of pigs fed formula in either the MEN period or during the full enteral feeding period. Pigs remained healthy only when colostrum was fed both during MEN and full EN (CC pigs), which further enhanced digestive enzyme activities, sugar-absorptive capacities, and gut structural parameters. The beneficial effects of MEN with colostrum are supported by our previous observations, where a more abrupt transition from PN to full enteral colostrum feeding was associated with NEC incidences of 15% to 30% and degrees of mucosal atrophy and dysfunction(n = 66).4

Minimal Enteral Nutrition and Necrotizing Enterocolitis / Cilieborg et al   39

T-RFLP: Dominating T-RFs (base pairs) 0.20

A

a a a,b

0.15 b

0.10

Relative intensities

0.05 0.00

b,c a

a,c a

61

217

219

223

231

233

OF 0.20

b

b

b

251

a

a,b a a

a

254

266

FF

347

CF

372

500

570

584

CC

B *

0.15 0.10 0.05 0.00 61

217

219

223

231

233

251

254

NEC

266

347

372

500

570

584

Healthy

Figure 5.   Microbiota in the distal small intestine mucosa investigated by terminal restriction fragment length polymorphism (T-RFLP). The diagram shows dominating terminal restriction fragments (T-RFs) (mean ± standard error of the mean). (A) Among the 4 treatments, means not sharing the same superscript letter differ significantly. (B) Between healthy and diseased (necrotizing enterocolitis [NEC]) piglets, only T-RF 584 differed in intensity (*, P < .05). CC, group receiving colostrum in both periods; CF, group receiving colostrum minimal enteral nutrition followed by full enteral formula feeding; FF, group receiving full formula in both periods; OF, control group receiving oral saline during the PN period and formula during the full enteral feeding period.

A

B

E

4

* C

D

FISH score

3

2

1

0 OF

FF

CF

CC

Figure 6.   Fluorescence in situ hybridization (FISH) analyses applied to cross-sections of the distal small intestine. Scanned photos showing bacteria (red signal) of representative pigs: OF (A), FF (B), CF (C), and CC (D). (E) FISH scores (mean ± standard error of the mean) differed significantly between OF pigs and the other 3 groups (P < .05). CC, group receiving colostrum in both periods; CF, group receiving colostrum minimal enteral nutrition followed by full enteral formula feeding; FF, group receiving formula in both periods; OF, control group receiving oral saline during the PN period and formula during the full enteral feeding period.

40   Journal of Parenteral and Enteral Nutrition / Vol. 35, No. 1, January 2011

80.0

SCFA concentrations

a

a

mmol/kg

60.0

a

a,b a,b

40.0

b

b,c

20.0

c

0.0 OF

FF

CF

CC

OF

Others

Succinate

FF

CF

CC

Colon contents

Stomach contents Acetate

Octanoate

Lactate

Figure 7.  Concentration of total and dominating short-chain fatty acids (SCFAs) in stomach and colon contents (total SCFAs: mean ± standard error of the mean). Means not sharing the same superscript letter differ significantly (P < .05). CC, group receiving colostrum in both periods; CF, group receiving colostrum minimal enteral nutrition followed by full enteral formula feeding; FF, group receiving formula in both periods; OF, control group control group receiving oral saline during the PN period and formula during the full enteral feeding period.

NEC incidence is lower among preterm infants (~10%) than in preterm pigs, which to a large degree can be explained by animal model considerations. Clinically relevant volumes of formula (120 mL/kg/d) induce spontaneous NEC in preterm pigs with incidences of ~40% if enteral feeding is initiated immediately after birth and ~70% if enteral feeding is initiated after a short PN period. This allows clinically relevant nutrition inventions using a limited number of animals. We have repeatedly shown that, in contrast to formula feeding, colostrum feeding immediately from birth results in a very low (< 10%) incidence of NEC in preterm pigs 4,26 Given the low incidences of NEC in infants, it is seldom a reported outcome for MEN interventions. A few studies and meta-analyses claim the safety and possible NEC-protective effects of small oral feeding doses but associate rapid advancement in feeding volume with increased NEC incidences.34-37 We show here that diet composition is an important factor in such studies. Our results suggest that MEN feeding with either diet type does not markedly alter NEC incidence and functional parameters if it is followed by full formula feeding. NEC, in contrast, is fully prevented in preterm pigs if colostrum MEN is followed by full colostrum feeding. However, lesions in the CF group were mainly restricted to the colon, which suggested that colostrum MEN provided some protection for the small intestine, but the feeding volume was likely too small to reach and affect the colon, which remained sensitive to inflammatory processes induced by later formula feeding. Formula feeding during MEN did not induce gut maturation or

NEC resistance, and we observed increased incidence of NEC and severe lesions throughout the small intestine in both OF and FF pigs. In the present study, 16 to 24 mL/kg/d as the MEN volume only had limited effects on gut structure. Only the small intestine from CC pigs maintained a high muscular tone (small intestine diameter) and higher villi, indicating that a colostrum diet is required throughout the neonatal period to protect against NEC lesions, mucosal atrophy, and loss of muscular tone. MEN volumes were probably too small to induce trophic effects on the intestine, which is consistent with studies in older pigs and young dogs that have shown enteral intake above 30% to 40% of daily energy needs was required to stimulate gut growth.5,6 However, MEN feeding may have induced small functional improvements, as indicated by the general increases in the enzyme activities in the proximal and middle intestine of MEN groups relative to OF control pigs. The marginal effect of MEN in this study was overshadowed by more pronounced diet and region-dependent effects of full enteral feeding, as indicated by the higher enzyme activities in CC pigs relative to CF pigs in the distal small intestine (except for sucrase). In this and our previous studies, lower sucrase activities have been correlated with a healthier epithelium, probably because of more differentiated enterocytes bearing less sucrase and more lactase activity.4,26 The limited effects of low MEN doses were supported by a later study with high colostrum MEN doses (48–108 mL/kg/d during a 3-day PN period), which resulted in increased small intestine weight and enzyme activities.38 Paradoxically, incidence of NEC was high also

Minimal Enteral Nutrition and Necrotizing Enterocolitis / Cilieborg et al   41

in the MEN group (66% vs 60% in OF control pigs), and although mainly ascribed to a high prevalence of stomach lesions,38 these results support the adverse findings in a human infant study with rapidly advanced enteral feeding.35 In infants, early MEN feeding (20 mL/kg/d starting day 4) vs late MEN feeding (starting day 15) resulted in increased in vivo lactase activity on day 10 and 28, with the highest levels in infants receiving mother’s milk relative to formula.39 MEN feeding may improve enterocyte function both directly and indirectly via reduced tissue inflammation, as suggested by an earlier study with preterm pigs in which intestinal hexose uptake in vitro was decreased in tissue sections that appeared healthy but originated from pigs with NEC lesions in an adjacent region.27 Formula feeding, both MEN and full enteral feeding, induced a significant decrease in plasma galactose and mannitol levels and may indicate that both the passive concentration-dependent diffusion (mannitol) and the active carrier-mediated transport by the sodium– glucose co transporter 1 (mediating glucose and galactose uptake) was affected by formula feeding. In addition, diet-dependent differences in gastric emptying, physical structure, availability of nutrients, and intestinal motility may be important. The microbiology of gastric contents showed that anaerobic bacteria were present in large amounts (107 CFU/mL) even before full enteral feeding on day 3. The stomach normally contains few bacteria, but in preterm newborns with a poorly developed acid barrier (pH 4–4.5), the gastric contents may play a key role in the initial colonization of the whole GI tract. Surprisingly, MEN feeding did not significantly affect stomach bacterial density or the pH of gastric aspirates, which suggested that the initial bacterial colonization in this region is largely independent of the availability of luminal substrates and the nature of the diet. This was confirmed by the T-RFLP analyses on small intestine tissue, which showed limited differences among the 4 treatment groups. Likewise, there were minimal differences in mucosal microbiota diversity between healthy and diseased pigs. This conflicts with several previous studies where C perfringens (represented by T-RF 233) was highly abundant in pigs with NEC,4,26,29 but this finding supports a recent preterm pig study that questions the causative role of C perfringens in NEC pathogenesis.40 However, FISH analyses of mucosal cross-sections suggested that OF pigs not fed MEN had significantly more bacteria in the distal small intestine than other groups. Concentrations of bacterial fermentation products (SCFAs) in the stomach and colon contents confirmed this finding, given that values were highest in OF pigs, intermediate in the 2 MEN formula groups (FF, CF), and lowest in the MEN colostrum group (CC). We

speculate that bacterial density and diversity in the preterm gut are not consistently related to the nature and amount of the initial enteral diet, but rather may be affected by the progression of inflammatory lesions resulting in bacterial overgrowth. Another recent preterm pig study indicated that colonization processes in preterm individuals is more random than in term individuals and that host responses controlling gut colonization are underdeveloped.40 It remains a challenge to feed the vulnerable preterm infant, and widely different strategies for reaching full enteral feeding exist. The results of the present study, coupled with those in previous studies of preterm pigs, suggest that feeding of the preterm neonate immediately after birth is not associated with a greater risk of NEC, regardless of diet. However, the risks of NEC and intestinal atrophy and dysfunction are much higher after formula feeding than after colostrum feeding. Supplementing PN with small doses of MEN (16–24 mL/kg/d) while delaying full enteral formula feeding marginally improved NEC resistance, intestinal dysfunction, and bacterial overgrowth, but the doses used were likely insufficient to provide a consistent beneficial response to later formula feeding. NEC and intestinal dysfunction were only absent when colostrum MEN was followed by full colostrum feeding. We conclude that colostrum improves intestinal health and maturation, but the doses often used for MEN in preterm infants are likely too small to provide significant beneficial effects against later formula-induced inflammation and malfunction.

Acknowledgments We thank Mette Schmidt, Richard Siggers, Hanne Møller, Søren Rasch, Chantal Lau, and Elin Skytte for assistance with the animal procedures. We also thank Annie Ravn Pedersen, Sophia Rasmussen, Joanna Amenuvor, Thomas Boye Pihl, and Katja Kristensen at the National Veterinary Institute, DTU, for assistance with microbiological analyses.

References 1. Sangild PT. Gut responses to enteral nutrition in preterm infants and animals. Exp Biol Med. 2006;231:1695-1711. 2. Neu J. Gastrointestinal development and meeting the nutritional needs of premature infants. Am J Clin Nutr. 2007;85(suppl):629S634S. 3. Diehl-Jones WL, Askin DF. Nutritional modulation of neonatal outcomes. AACN Clin Issues. 2004;15:83-96. 4. Bjornvad CR, Thymann T, Deutz NE, et al. Enteral feeding induces diet-dependent mucosal dysfunction, bacterial proliferation, and necrotizing enterocolitis in preterm pigs on parenteral nutrition. Am J Physiol Gastrointest Liver Physiol. 2008;295: G1092-G1103.

42   Journal of Parenteral and Enteral Nutrition / Vol. 35, No. 1, January 2011

5. Burrin DG, Stoll B, Jiang R, et al. Minimal enteral nutrient requirements for intestinal growth in neonatal piglets: how much is enough? Am J Clin Nutr. 2000;71:1603-1610. 6. Owens L, Burrin DG, Berseth CL. Minimal enteral feeding induces maturation of intestinal motor function but not mucosal growth in neonatal dogs. J Nutr. 2002;132:2717-2722. 7. Niinikoski H, Stoll B, Guan X, et al. Onset of small intestinal atrophy is associated with reduced intestinal blood flow in TPN-fed neonatal piglets. J Nutr. 2004;134:1467-1474. 8. Kansagra K, Stoll B, Rognerud C, et al. Total parenteral nutrition adversely affects gut barrier function in neonatal piglets. Am J Physiol Gastrointest Liver Physiol. 2003;285:G1162-G1170. 9. Burrin DG, Stoll B, Chang X, et al. Parenteral nutrition results in impaired lactose digestion and hexose absorption when enteral feeding is initiated in infant pigs. Am J Clin Nutr. 2003;78:461-470. 10. Harvey RB, Andrews K, Droleskey RE, et al. Qualitative and quantitative comparison of gut bacterial colonization in enterally and parenterally fed neonatal pigs. Curr Issues Intest Microbiol. 2006;7:61-64. 11. Simeunovic E, Arnold M, Sidler D, Moore SW. Liver abscess in neonates. Pediatr Surg Ing. 2009;25:153-156. 12. Duro D, Kamin D, Duggan C. Overview of pediatric short bowel syndrome. J Pediatr Gastroenterol Nutr. 2008;47 (suppl 1):S33-S36. 13. Kelly DA. Intestinal failure-associated liver disease: what do we know today? Gastroenterology. 2006;130 (suppl 1):S70-S77. 14. McGuire W, Bombel IS. Slow advancement of enteral feed volumes to prevent necrotising enterocolitis in very low birth weight infants. Cochrane Database Syst Rev. 2008;(2):CD001241. 15. Berseth CL. Feeding strategies and necrotizing enterocolitis. Curr Opin Pediatr. 2005;17:170-173. 16. Tyson JE, Kennedy KA. Trophic feedings for parenterally fed infants. Cochrane Database Syst Rev. 2008;(3):CD000504. 17. Duffy LC, Zielezny MA, Carrion V, et al. Bacterial toxins and enteral feeding of premature infants at risk of necrotizing enterocolitis. Am J Hum Biol. 1998;10:211-219. 18. Penders J, Thijs C, Vink C, et al. Factors influencing the composition of the intestinal microbiota in early infancy. Pediatrics. 2006;118:511-521. 19. Harmsen HJ, Wildeboer-Veloo AC, Raangs GC, et al. Analysis of intestinal flora development in breast-fed and formula-fed infants by using molecular identification and detection methods. J Pediatr Gastroenterol Nutr. 2000;30:61-67. 20. Schwiertz A, Gruhl B, Löbnitz M, Michel P, Radke M, Blaut M. Development of the intestinal bacterial composition in hospitalized preterm infants in comparison with breast-fed, full-term infants. Pediatr Res. 2003;54:393-399. 21. Wang Y, Hoenig JD, Malin KJ, et al. 16S rRNA gene-based analysis of fecal microbiota from preterm infants with and without necrotizing enterocolitis. ISME J. 2009:3:944-954. 22. de la Cochetiere MF, Piloquet H, des Robert C, Darmaun D, Galmiche JP, Roze JC. Early intestinal bacterial colonization and necrotizing enterocolitis in premature infants: the putative role of Clostridium. Pediatr Res. 2004;56:366-370. 23. Blakey JL, Lubitz L, Campbell NT, Gillam GL, Bishop RF, Barnes GL. Enteric colonization in sporadic neonatal necrotizing enterocolitis. J Pediatr Gastroenterol Nutr. 1985;4:591-595.

24. Hoy CM, Wood CM, Hawkey OM, Puntis JW. Duodenal microflora in very-low-birth-weight neonates and relation to necrotizing enterocolitis. J Clin Microbiol. 2000;38:4539-4547. 25. Millar MR, Linton CJ, Cade A, Glancy D, Hall M, Jalal H. Application of 16S rRNA gene PCR to study bowel flora of preterm infants with and without necrotizing enterocolitis. J Clin Microbiol. 1996;34:2506-2510. 26. Sangild PT, Siggers RH, Schmidt M, et al. Diet- and colonizationdependent intestinal dysfunction predisposes to necrotizing enterocolitis in preterm pigs. Gastroenterology. 2006;130:1776-1792. 27. Buddington RK, Bering SB, Thymann T, Sangild PT. Aldohexose malabsorption in preterm pigs is directly related to the severity of necrotizing enterocolitis. Pediatr Res. 2008;63:382-387. 28. de Man JC, Rogosa M, Sharpe E. A medium for the cultivation of Lactobacilli. J Appl Microbiol. 1960;23:130-135. 29. Siggers RH, Thymann T, Jensen BB, et al. Elective cesarean delivery affects gut maturation and delays microbial colonization but does not increase necrotizing enterocolitis in preterm pigs. Am J Physiol Regul Integr Comp Physiol. 2008;294:R929-R938. 30. Mølbak L, Thomsen LE, Jensen TK, Bach Knudsen KE, Boye M. Increased amount of Bifidobacterium thermacidophilum and Megasphaera elsdenii in the colonic microbiota of pigs fed a swine dysentery preventive diet containing chicory roots and sweet lupine. J Appl Microbiol. 2007;103:1853-1867. 31. Amann RI, Binder BJ, Olson RJ, Chisholm SW, Devereux R, Stahl DA. Combination of 16S ribosomal-RNA-targeted oligonucleotide probes with flow-cytometry for analyzing mixed microbialpopulations. Appl Environ Microbiol. 1990;56:1919-1925. 32. Jensen MT, Cox RP, Jensen BB. Microbial production of skatole in the hind gut of pigs given different diets and its relation to skatole deposition in backfat. Anim Sci. 1995;61:293-304. 33. Bombell S, McGuire W. Delayed introduction of progressive enteral feeds to prevent necrotizing enterocolitis in very low birth weight infants [review]. Cochrane Database Syst Rev. 2008;(2):CD001970. 34. Chauhan M, Henderson G, McGuire W. Enteral feeding for very low birth weight infants: reducing the risk of necrotizing enterocolitis. Arch Dis Child Fetal Neonatal Ed. 2008;93:F162-F166. 35. Berseth CL, Bisquera JA, Paje VU. Prolonging small feeding volumes early in life decreases the incidences of necrotizing enterocolitis in very low birth weight infants. Pediatrics. 2003;111:529-534. 36. Davey AM, Wagner CL, Cox C, Kending JW. Feeding premature infants while low umbilical artery catheters are in place: a prospective, randomized trial. J Pediatr. 1994;124:795-799. 37. Eyal F, Sagi E, Arad I, Avital A. Necrotising enterocolitis in the very low birthweight infant: expressed breast milk feeding compared with parenteral feeding. Arch Dis Child. 1982;57:274-276. 38. Vegge A, Thymann T, Sangild PT. Replacing lipids with glucose in parenteral nutrition increases susceptibility to necrotizing enterocolitis in preterm pigs [abstract]. Clin Nutr. 2009:4(suppl 2): 118. 39. Schulman RJ, Schanler RJ, Lau C, Heitkemper M, Ou CN, Smith EO. Early feeding, feeding tolerance, and lactase activity in preterm infants. J Pediatr. 1998;133:645-649. 40. Cilieborg MS, Boye M, Mølbak L, Jensen ML, Thymann T, Sangild P. Preterm birth and necrotizing enterocolitis alter gut colonization in pigs. Pediatr Res. 2010 (in press).