Validation of protocol of experimental necrotizing enterocolitis in rats and the pitfalls during the procedure

PURPOSE: To describe the difficulties of implementing the protocol of experimental necrotizing enterocolitis (NEC) in order to obtain a larger number of newborns affected with the disease and a lower mortality. METHODS: Term Sprague-Dawley newborns rats (22 days) were divided into four groups of 12 fetuses each (n = 48): EC – breastfed newborns; IH – breastfed newborns and subjected to a stress protocol by ischemia and hypothermia; ESB – formula-fed newborns (Esbilac®, PetAg, Hampshire, IL, USA) and NEC – formula-fed newborns and subjected to stress protocol. The parameters set for the study protocol were: milk concentration (0.19 g ml or 0.34 g/ml), diet instilled volume (according to body weight – 200 kcal/day/Kg – or progressive, according to acceptance), weight (gain, loss or maintenance) and duration of the experiment (72 hours or 96 hours). Data of body weight (BW), intestinal weight (IW) and the IW/BW ratio were obtained. Samples of terminal ileum were collected and analyzed by the degree of injury to the intestinal wall. Statistically significance was set to p<0.05. RESULTS: The established protocol with less mortality and increased number of NEC was with Esbilac® at a concentration of 0.19 g/ ml of diet instilled volume of 0.1 ml, every 3 hours, for 72 hours. All infants fed with artificial milk lost weight. In the degree score of intestinal injury, the ESB, IH and NEC groups were considered positive for NEC with greater histological injury in the latter. CONCLUSION: The described NEC protocol in rats allowed a greater survival of puppies with a greater number of animals affected by the disease.


Introduction
Necrotizing enterocolitis (NEC) is a severe inflammatory bowel disease that affects the intestines of neonates and in whose genesis is involved hyperosmolar feeding, infection and ischemia.
The disease is characterized by intestinal necrosis, sepsis and multiple organ failure 1 , with overall incidence of 1:1000 live births, reaching 2-5% of all preterm newborns 2,3 and represents the leading cause of death by gastrointestinal disease in the neonatal period.
The development of animal models using newborn pigs, mice and rats can be used to reliably reproduce the characteristics of this disease and represents one of the approaching ways of fundamental importance for understanding the pathogenesis of NEC.
The first description of the NEC model in newborn rats was reported by Barlow et al. 4 that induced the disease by combining formula feeding, followed by intermittent episodes of hypoxia stress, then, Barlow & Santulli added hypothermia to the model made for three days 5 .Caplan et al. modified the model by adding an additional day of hypothermia, from three to four days, and standardized ischemia using nitrogen 6 .Dvorak et al. used the same model with hypoxia for 4 days, the formulation of an own rat milk substitute and created a histopathologic grading to determine the extent of damage to the ileum 7 .In Brazil, Meyer et   al. evaluated in the NEC model in rats, just the degree of intestinal involvement with hypoxia and reperfusion without using formula or hypothermia 8 .
Because of the importance and the incidence of NEC in the neonatal intensive care units and the variety of ways to obtain the experimental disease in rats, as there are variations in time, of the milk formulation and its concentration, and way of ischemia, we decided to carry out an experimental protocol of NEC to establish what would be the proper way of inducing the disease and report technical difficulties of its implementation.Ischemia and hypothermia are well explained in other studies, however the procedure for milk administration is not well described.Thus, the following parameters were established to the formation of Esbilac ® (PetAg, Hampshire, IL, USA) formula group: milk concentration (0.19 g/ml or 0.34 g/ml), volume of instilled diet (according to body weight -200 kcal/day/Kg -or progressive, according to acceptance), harvest body weight (gain, loss or maintenance, in g, compared to birth weight) and duration of the experiment (72 or 96 hours).

Formation of the experimental groups after standardization of the protocol
On day 22 of gestation, the rats were weighed and labor was induced by injection of 1 IU of oxytocin (Oxiton ® , União Química, Brazil) subcutaneously.All puppies were weighed at birth and randomly directed to four major groups of 12 fetuses each, performing a total of 48 fetuses.The groups established for the new protocol were: 1) External control group (EC): neonates who did not undergo manipulation and were breastfed by its mother; 2) Ischemia/hypothermia group (IH): neonates who were exclusively by their mother, but were subjected to hypoxia, exposing pups to 100% nitrogen for 60 seconds, followed by hypothermia, by exposure to cold (4°C) for 10 minutes, twice a day according to Caplan et al. 6 ; 3) Esbilac group (ESB): the newborns were fed with artificial milk (Esbilac ® ), instilling 0.1 ml every 3 hours through a BD™ PICC catheter (26G -1.9 Fr -Single Lumen PICC, 50 cm) for orogastric feeding in the first 24 hours, progressively increasing the volume to 0.2 ml, if tolerated; and 4) Necrotizing enterocolitis group (NEC): consistent of protocols of the groups ESB and HI, where the newborns were formula fed and underwent ischemia, and hypothermia.Immediately after birth, neonates of the groups ESB and NEC were numbered and placed individually in a compartmentalized box with sawdust over a heated table (Harvard Apparatus®), previously set to 38°C (Figure 1).

Evaluation of technical difficulties
The parameters were: gastric distension, bronchial aspiration (confirmed at necropsy), achievement of neonates with NEC and survival.

Morphological evaluation of groups
Body weight (BW) and intestinal weight (IW) were measured in grams (g) and the ratio IW/BW was performed in order to exclude the variable BW on the evaluation of IW.

Processing for histological analysis
A 2 cm segment of ileum, immediately proximal to the ileocecal valve was removed for analysis.After fixation in 4% paraformaldehyde, the samples were dehydrated in an increasing gradient of ethanol, cleared in xylene and embedded in histology paraffin.Histologic sections of 5μm thickness were performed and stained with Ehrlich hematoxylin/eosin (H/E).

Histological grading of necrotizing enterocolitis
Each slice of each lamina was analyzed according to the graduation described by Dvorak et al. 9 in: 0: no damage; 1: slight submucosal and/or lamina propria separation; 2: moderate submucosal and/or lamina propria separation, and/or edema in submucosal and muscular layers; 3: severe separation of submucosa and/or lamina propria, and/or severe edema in submucosa and muscular layers, region villous sloughing; 4: loss of villi and necrosis.Intermediate scores of 0.5, 1.5, 2.5 and 3.5 were also used for a more accurate evaluation of the levels of intestinal damage.To determine the incidence of NEC, animals with histological scores < 2 did not have NEC and those with histological score ≥ 2 had NEC.

Statistical analisys
The obtained values of weight and histological grading were evaluated using nonparametric ANOVA with Tukey-Kramer post-test, considering significant differences with a p<0.05.
The results were expressed as mean ± standard deviation (SD).
Calculations were made using GraphPad Prism 3.02.

Evaluation of technical difficulties to adaptation of protocol
The diet with hyperosmolar milk concentrated to 0.34 g/ ml presented 77% of bronchial aspiration and 97% of mortality.
Reducing the concentration to 0.19 g/ml led to a decrease of bronchial aspiration to 19% and mortality to 48% (Table 1).The initial instilled volume was from 0.1 ml to 0.25 ml using as a according to the weight of the newborn (200 kcal/day/ Kg).Subsequently the volume of instillation was changed to 0.1 ml in the first 24 hours, reaching up to 0.2 ml in the next 48 hours, using gastric distension as a parameter to volume progression.
There was a progressive weight loss in both, 72 hours and 96 hours, protocols.However, the mortality rate was lower in the 72 hours protocol than in the 96 hours (respectively 21% and 97%) (Table 2).The intestinal weight (IW) showed a significant difference between the EC and IH groups compared to the ESB and NEC groups (p<0.001).The IH and EC groups compared with each other showed no significant differences, as well as the ESB and NEC groups (p>0.05).
As for IW / BW ratio, only significant difference observed was in comparison of EC group with IH and ESB groups (p<0.05)(Table 3 and Figure 2).

NEC evaluation
The general macroscopic and microscopic aspects were: All animals with scores ≥ 2 were considered positive for the NEC and this occurred in the ESB, IH and NEC groups for most newborns harvested after 72 hours of life, however, the scores of NEC puppies were the highest of all when compared with EC (p<0.001) (Table 3 and Figure 3).

Discussion
In the experimental NEC model in rats fed by using milk formula associated with asphyxia and hypothermia is possible to obtain 80% of newborns with the disease 7 .This model includes the two most important risk factors for NEC that are immaturity and enteral feeding with formula, since the development of many essential intestinal functions in rats occurs in the early postnatal period, its use is similar to the essential functions that also occurs in the human immature intestine of newborns 10 .
Asphyxia and hypothermia are well established procedures in this model using, respectively, nitrogen and followed by cooling to 4°C for 10 minutes, yet the procedure for milk administration is very variable.Therefore, in the formation of the ESB group (Esbilac ® milk) there are four variables to be considered: milk concentration, instilled volume of diet, body weight of the newborn and duration of the experiment to obtain the NEC (ranging from 72 or 96 hours).
The feeding of newborns is very time-consuming and laborious, and you need several members to perform it.We tried to feed them with a 24G Teflon catheter (Abbocath ® ), but we had some cases of oral bleeding and even oropharygeal perforation with subsequent death.The success rate of only increased by the use of a 1.9Fr silicone catheter PICC type, it is more delicate and less rigid than the Teflon catheter.
The Esbilac ® administered volume was initially based on the model of Caplan et al. 6 which administered from 0.1ml to 0.4ml (200 kcal/kg/d), with 3-hour interval between each diet, for a period of 96 hours.
In order to obtain an hyperosmolar diet, we prepared a concentrated solution of Esbilac ® milk (0.34g/ml) diluted in distilled water, twice the concentration (2:1) recommended by the manufacturer.However, there was high incidence of bronchial aspiration and high mortality.We, therefore, decided to lower the concentration to 0.19g/ml (1:2), resulting in decrease of bronchial aspiration and easier instillation of the diet, since in the concentrated solution milk appearance was very viscous.We realized, even after the change in milk concentration, and despite the lower rates of aspiration, that the mortality of newborns only with the milk was still high, especially in the third and fourth days of life.
Then we evaluate the volume of administered diet to newborns.The initial calculation for the volume of infusion was carried out according to infant weight varying from 0.1 ml to 0.4 ml as tolerated.We expected that the animals gained weight, but they gradually lost it.These results are similar to Dvorak et al. 7 and other authors 11,12 who believe that the instilled milk volume can interfere with the digestive process since the infant does not digest the whole milk throughout the digestive tract.

Conclusion
The established protocol with larger number of survivors and larger number of NEC was with Esbilac ® , at a concentration of 0.19 g/ml, instilled volume of diet of 1 ml every 3 hours, when no gastric distension was observed, during 72 hours.
Animal Experimentation All procedures involving animals were in accordance with the guidelines established by COBEA (Brazilian College on Animal Experimentation) and approved by the Ethics Committee on Animal Experimentation of School of Medicine of Ribeirao Preto (CETEA -School of Medicine of Ribeirao Preto, USP) under number 040/2011.Animals and pregnant females achievement Male and female adult Sprague-Dawley rats from the Multidisciplinary Center for Biological Research (CEMIB -UNICAMP, Campinas, SP, Brazil) were submitted to mating during the dark cycle.On the following day, the female genital region was examined and a vaginal smear was performed for observation of sperm.The presence of sperm configured mating and was considered day zero of pregnancy (term = 22 days).The animals were kept in cages, with water and food for rodents ad libitum, under controlled lighting (12 hours of light/12 hours of darkness), temperature (mean of 23 °C) and relative air humidity (mean of 55%) conditions.Validation of protocol of experimental necrotizing enterocolitis in rats and the pitfalls during the procedure Acta Cirúrgica Brasileira -Vol.28 (supl.1) 2013 -21 Evaluation of technical difficulties to adaptation of protocol To perform the NEC protocol three procedures are essential: ischemia, hypothermia and artificial milk administration.

FIGURE 1 -
FIGURE 1 -Methodology used for administration of artificial milk and NEC protocol.A) Compartmentalized box with sawdust for puppies; B) Observation if gastric distension (arrow) surpasses or not the midline before feeding the newborn; C) Orogastric catheterization with 1.9 Fr PICC catheter; D) Pups after hypoxia with N 2 100% for 60 seconds.
a) the EC group presented a healthy gut, intact intestinal villi and layers, b) the IH group presented a darkened bowel and the intestine with a more severe separation of the lamina propria from the submucosa, besides villous lesions; c) the ESB group presented an intestine with a swollen aspect, and some with inflammatory aspect, histologically with moderate to severe separation of the lamina propria from the submucosa and d) the NEC group presented a very dark and swollen intestine, and in most of the neonates the loops were necrotic and histologically with partial or total loss of villi.

FIGURE 3 -
FIGURE 3 -Morphology and histology of the ileum of the four studied groups.A and E) EC group -villous of normal histological pattern; B and F) ESB group -moderate injuries; C and G) IH group -moderate to severe injuries; D and H) NEC group -severe lesion with destruction of intestinal villi.

TABLE 1 -
Mortality and aspiration rates according to the concentration of administered milk.

TABLE 2 -
Mortality rate in the harvest according to

TABLE 3 -
Graphics comparing the harvest BW, IW, IW/BW ratio and NEC scores per group.
. Mean and standard deviation for the four studied groups EC, IH, ESB and NEC: evolution of body weight (BW), intestinal weight (IW), IW/BW ratio,% of death rate (DR) between the groups and NEC scores per group.* p<0.05; # p<0.001 a = Compared with Ischemia/Hypothermia (IH); b = Compared with Esbilac (ESB); c = Compared with Necrotizing Enterocolitis (NEC).
The four developed groups (EC, IH, ESB, NEC) showed differences in morphological analysis.The intestinal weight of the breastfed groups was higher than of those fed with artificial milk.However, the IW/BW ratio was increased in IH and ESB groups probably due to by injury and subsequent inflammation of this organ.The same result could have happened to the NEC group, but the IW/BW ratio was not different from the EC group, this may have happened due to the necrotic process of intestinal cells causing tissue decrease, what is corroborated by the results found by histological evaluation.Our histological results showed that breast-feeding preserved intestinal structure, this protection, though fundamental, was not sufficient to prevent ischemic injury in the organ as demonstrated by the histological grade of moderate and severe injury in neonates of IH group.Similarly, the histological results of ESB group showed moderate to severe effect, showing that only with the Esbilac ® milk diet may be sufficient to affect the integrity of intestinal epithelium.Finally, in the NEC group, histological changes were more severe, with flattening or local absence of villi, characteristic of tissue necrosis.