1. Introduction

Necrotizing Enterocolitis (NEC), a common gastrointestinal complication in newborn infants, is characterized by variable damage to the intestinal tract, ranging from mucosal injury to full-thickness necrosis and perforation. NEC occurs in 1 to 3 per 1000 live births and 1 to 7.7 percent of admissions to neonatal intensive care units. Although early recognition and aggressive treatment of this disorder has improved clinical outcomes, NEC accounts for substantial long-term morbidity in survivors of neonatal intensive care, particularly in premature very low birth weight (BW) infants (<1500 g). The mortality rate (15%-25%) for affected infants has not changed appreciably in 30 years [1,2].

The optimal treatment of NEC, via medical intervention, peritoneal drainage, or laparotomy is also controversial. Many infants with NEC recover uneventfully with medical therapy and have long-term outcomes similar to unaffected infants of matched gestational age. Infants with progressive disease requiring peritoneal drainage and/or surgical intervention suffer almost all of the mortality and morbidity. Of these, approximately 30%-40% will die of their disease and most of the remainder will develop long-term neurodevelopmental and gastrointestinal morbidity [2].

Current work is focusing on developing a better understanding of the pathogenesis and improving means to identify which infants are at greatest risk of disease progression and complications. This includes, post-NEC intestinal strictures, which affect up to one third of patients and are a major driver of severe prolonged morbidity. Intestinal strictures promote bacterial overgrowth in the small bowel, often leading to repeated infections, bloody stools, failure to thrive, and bowel obstruction [3]. Although the incidence of strictures is high, its risk factors are not well studied. One prospective observational study reported abnormal values of C-Reactive Protein (CRP) in both stage II and stage III NEC. CRP returned to normal at a mean of 9 days after initiation of appropriate medical management except in those who developed complications such as stricture or abscess formation [4]. In another retrospective study, the mean maximum CRP concentration during acute phase was significantly higher in infants who developed stricture (p<0.001), as was the mean duration of the elevation of CRP levels (p<0.001) [5].Various other laboratory parameters, including full blood count, platelet count, duration of thrombocytopenia and pH have been studied, however, none have consistently been shown to predict the development of post-NEC strictures[5-11].There is also a lack of studies on the impact of stricture location (i.e. small versus large bowel)on the time to presentation and diagnosis of post-NEC intestinal obstruction.

In this study, we aimed to define the incidence and location of post-NEC strictures in infants who underwent peritoneal drainage, laparotomy or medical management at a single major pediatric teaching hospital. We also evaluated potential predictors of post-NEC stricture development and explored the time to presentation and diagnosis of strictures in infants with NEC.

2. Methods

We performed a retrospective study of infants with a diagnosis of NEC who were treated at the Hospital for Sick Children (Toronto, Ontario, Canada) from 2003-2013. Infants who died within the first month of life were excluded.

We conducted a comprehensive chart review and collected demographic data, including BW and gestational age (GA). Clinical data including the presence and location of post-NEC strictures, delay in presentation and diagnosis of stricture (days), type of treatment (medical management, peritoneal drainage, surgery), and Bell’s NEC stage (I, IIA, IIB, III) was also obtained. Infants with stage I NEC, or suspected disease, were excluded from analysis. Stage IIA and IIB NEC was defined as mild and moderate disease respectively. Infants with stage IIA NEC presented with systemic signs including apnea, bradycardia and lethargy, in addition to grossly bloody stool, absent bowel sounds, and pneumatosis intestinal is on radiographs. Stage IIB was diagnosed if an infant presented with the abovementioned signs and symptoms, plus definite abdominal tenderness, right lower quadrant mass and/or ascites. Infants with Stage III NEC generally presented with all of the above in addition to severe systemic deterioration consisting of marked acidosis, hypotension, apnea, disseminated intravascular coagulation and neutropenia. Medical treatment generally consisted of feeding cessation with broad-spectrum antibiotics, the maintenance of vital hemodynamic and respiratory function, analgesia and parenteral nutrition. Peritoneal drainage, performed under local anesthesia, involved the placement of a drainage catheter in the abdominal cavity. Finally, the surgical protocol used involved exploratory laparotomy with detailed examination of the entire small and large bowel and identification of perforations, excision of non-viable bowel with primary anastomosis or stoma creation. We also collected laboratory data including maximum CRP levels, minimum platelet count, duration of thrombocytopenia (days), and ph.

Univariate analyses were then completed to assess for associations between the independent variables described above and stricture development. The results of our univariate analysis were used to identify variables to be studied using a multivariate model. This included birth weight, intervention, NEC stage, maximum CRP values, duration of thrombocytopenia and blood ph.GA and minimum platelet count were found to be co-linear with BW and duration of thrombocytopenia respectively, and thus, were excluded from the multivariate model.

Our secondary analysis consisted of evaluating differences in location of stricture and its impact on time to presentation of symptoms of obstruction and diagnosis of stricture. Time to presentation of symptoms was defined as the time period between cessation of treatment for NEC and onset of symptoms such as distention, vomiting, and obstipation. Time to diagnosis was defined as the period between onset of symptoms to confirmation of post-NEC stricture on contrast study. Since time to presentation of symptoms and diagnosis were both non-normally distributed variables, we used median values and non-parametric Wilcoxon-Rank Sum testing for our comparative analysis.

For all analyses, p-values ≤0.05 were considered significant. All statistical analyses were conducted with SPSS (IBM Corp. Released 2013. IBM SPSS Statistics for Windows, Version 22.0. Armonk, NY: IBM Corp.). The study protocol was approved by the Hospital for Sick Children Research Ethics Board prior to study initiation.

3. Results

A total of 175 infants, 94 males and 81 females, with NEC were included in the primary analysis. Table 1 details the demographic characteristics of the patient cohort. The mean GA and BW was 31.8 weeks and 1743.8 grams respectively. A total of 87 (50%), 65 (37%) and, 18 (10%) patients were treated with medical management, laparotomy and, peritoneal drain respectively. Nearly half (51%) of the population had Stage IIA disease at the time of diagnosis. A total of 14 (8%) and 72 (41%) infants were diagnosis with stage IIB and stage III disease, respectively (Table 1).

35 (20%) patients developed post-NEC strictures (Figure 1). Univariate analysis revealed several predictors of post-NEC stricture development (Table 2). Infants treated surgically, with laparotomy and resection, were significantly more likely to develop intestinal strictures (p<0.01). A majority of these patients (63%) underwent resection and primary anastomosis, of which about 57% developed strictures at the anastomotic suture line. Likewise, infants with higher NEC stage (p=0.013), CRP elevation (p<0.01), greater duration of thrombocytopenia (p=0.011), lower platelet counts (p=0.018), and academia(p=0.028) were at higher risk of developing post-NEC strictures (Table 2). After multivariate analysis, however, only elevation in CRP was shown to be predictive of stricture occurrence (p<0.047; OR 1.01; Table 3).

Notably, patients with strictures in the small bowel took significantly longer to present with symptoms of obstruction after cessation of therapy, compared to those with colonic strictures (35 days vs. 19 days; p=0.037; (Table 4A). The former cohort also experienced a relative delay (21 days vs. 12 days) in definitive diagnosis after onset of symptoms. This difference however, did not achieve statistical significance (p=0.09; Table 4B). 

4. Discussion

This retrospective chart review reports the incidence, risk factors and time to presentation and diagnosis of post-NEC strictures in a cohort of 175 infants treated with medical management, peritoneal drainage or laparotomy. Our initial analysis revealed 35 infants (20%) with post-NEC strictures. After multivariate analysis, solely the CRP value was found to be an independent predictor of post-NEC stricture development. For everyone increase in CRP, there was a 1.01 increased odds of post-NEC stricture development. Infants with small bowel strictures also experienced a significantly longer median time to presentation of symptoms of obstruction (35 days) than those with strictures in the large bowel (18.5 days). While there was a trend towards delay in diagnosis, this difference was not statistically significant. 

We report a 20% rate of post-NEC strictures in the overall study population. This rate is relatively lower than that reported in the recent literature [5,7,12]. For instance, Gaudin et al. reviewed 60 cases of NEC at a single tertiary center and reported a stricture rate of up to 57% [5]. It is important to note that in most of these studies, contrast study was routinely performed to screen for strictures. At our institution, however, patients are only investigated upon presenting with symptoms of obstruction or stricture. Screening for strictures in asymptomatic patients can increase reported rate of strictures from 17% to 36% [13] and may explain the difference in incidence of strictures observed in other reports compared to ours. Though screening with contrast study might prevent complications such as septicemia, perforation, and severe intestinal obstruction, we believe that this practice exposes infants to unnecessary radiation since most asymptomatic patients with strictures do not require intervention. In fact, systemic screening with contrast study is often performed early in the disease process. Histological evidence suggests that this practice is more likely to reveal strictures as it detects inflammatory lesions that tend to regress naturally [13]. Therefore, some suggest contrast study should only be performed at or after 6 weeks, followed by cautious re-feeding [14-16].

Weobserved a similar rate of stricture in NEC cases treated with laparotomy (34%) in comparison to other published reports [12]. As far as surgical treatment is concerned, the techniques used are numerous and vary from center to center [17, 18]. In our hospital, the surgical treatment of choice in most cases was laparotomy with resection of the necrotic zones and primary anastomosis. Necrotic regions were left un-resected in palliative patients with end-stage disease or in advanced cases that warranted a planned second look surgery after allowing the bowel to demarcate with hopes of minimizing the amount of bowel requiring resection. Resection during the acute phase allows the digestive tract to be better conserved since the disappearance of necrotic lesions leads to a decrease in the inflammatory and infectious phenomena that promotes the spreading of lesions [5]. This technique however, may promote the formation of intestinal strictures due to scarring and subsequent narrowing at the anastomotic suture line. Alternatively, a primary anastomosis may leak, leading to further inflammation and infection in the region causing the area to heal with an intestinal stricture. This phenomenon is reflected in our patient cohort, as patients treated with laparotomy and primary anastomosis were at increased risk of developing intestinal strictures, nearly 60% (8/14) of which occurred at the anastomotic suture line. Some authors propose proximal diversion, without resection of necrotic regions of the bowel. While this technique is thought to reduce the extent of resection and to increase the final length of the digestive tract, it may necessitate secondary resection of necrotic zones that undergo stricture at a later time [5].

One of our primary objectives was to study the risk factors that may predict the development of post-NEC strictures. In our study, having advanced disease (i.e. stage III) correlated significantly with the occurrence of post-NEC strictures. Our findings are consistent with other reports in which infants with Bell’s stage I or II disease as well as non-specific intestinal dilatation were significantly less likely to develop a stricture [12]. In addition to being a marker of extensive disease, a high Bell’s stage also represents a heightened immune mediator response and therefore, is a biologically plausible risk factor for the development of intestinal strictures. Similarly, a marked elevation of CRP levels, ≥49.5 mg/L, was significantly correlated with the occurrence of post-NEC strictures. After multivariate analysis, CRP elevation was the only variable that predicted the risk of post-NEC strictures in our patient cohort. These findings correlate with other reports of a potential relationship between the severity of the inflammatory syndrome and the chance of developing an intestinal stricture. A recent retrospective study demonstrated that the mean maximum CRP concentration and the mean duration of CRP elevation was significantly higher in infants who developed post-NEC strictures [5]. Likewise, in their prospective observational study, Pourcyrous et al. concluded that in infants with suspected NEC, normal CRP values should prompt abortion of antibiotic therapy and resumption of feeds. Furthermore, they observed that the CRP values of infants with stage II NEC returned to normal at a mean of 9 days except in those who developed complications including stricture formation. Persistent elevation of CRP levels, therefore, was related to the occurrence of post-NEC strictures requiring surgical intervention [4]. Our results confirm the prognostic relevance of CRP, a sensitive marker of systemic inflammation, to the risk of developing post-NEC intestinal strictures. In the future, one may also investigate the utility of other markers of inflammation, such as fecal calprotectin, interleukin-6 and pro-calcitonin in the development of post-NEC strictures [19].

Furthermore, in our cohort of175 infants, we identified 35 infants (20%) with small bowel strictures and 33 infants (19%) with large bowel strictures. In contrast to these findings, several studies in the literature have reported a relatively higher incidence of strictures in the colon than in the small bowel [5,12,20-23]. The most common site of post-NEC strictures is thought to be the left colon, especially around the vascular watershed region of the splenic flexure [23]. However, our findings indicate that in addition to colonic strictures, physicians and surgeons must also maintain a high index of suspicion for strictures in the small intestine. In our study, infants with small bowel structures took significantly longer to present with symptoms of obstruction after cessation of treatment than those with structures in the large bowel. This delay in presentation was accompanied by a subsequent delay in establishing a definitive diagnosis in this subgroup of patients. To our knowledge, this is the first study to evaluate the difference in time to presentation and diagnosis of small and large bowel post-NEC strictures. We postulate that the difference in diagnosis is likely due to the inability of the contrast medium to penetrate the far reaches of small bowel, especially in areas near the mid-small bowel and proximal ileum. Re-absorption of the contrast medium prior to reaching these areas can lead to missed and/or delayed diagnoses, which further complicates the care of infants with NEC. There have been reports of patients with intestinal obstruction from small bowel strictures despite normal contrast study [13]. One may mitigate this problem with innovative imaging modalities, such as the ones being used in the management of structuring Crohn’s disease [24,25]. Future studies must be aimed at studying their clinical utility in comparison to traditional imaging options such as contrast enema and small bowel follow through, in the timely diagnosis and treatment of small bowel strictures. 

The results of our study are limited by the retrospective, single center study design. Regular monitoring of CRP and other laboratory markers also did not become common practice at our center until 2013, leaving a large proportion of patients with limited data and eventual exclusion from the study. There may be additional risk factors, such as the degree of leukocytosis or length of resected bowel [12], for the development of post-NEC strictures that were present in our patient population but not studied. Lastly, ours is a tertiary/ quaternary pediatric facility with referrals from other neonatal units for surgical evaluation. These cases may represent the severe end of the NEC spectrum; our results therefore, may not apply to other perinatal settings that manage infants with milder cases of NEC. 

Physicians must maintain a lower threshold for the development of intestinal strictures in patients with advanced NEC, elevated inflammatory markers and need for surgical intervention. Prospective studies are required to validate these risk factors and develop a predictive model for post-NEC strictures. Our findings also suggest that obstructive symptoms from small bowel strictures may take longer to manifest clinically; this warrants further study of innovative imaging modalities to enable timely diagnosis of strictures in this vulnerable patient population.

5. Sources of Funding

None. 

6. Conflicts of Interest 

None. 

This work was accepted for an oral presentation at the European Paediatric Surgeons’ Association Meeting in June 2016 (Milan, Italy).

Table 1: Baseline Characteristics.

Table 2: Univariate Predictors of Stricture Development.

Table 3: Multivariate Predictors of Stricture Development.

Table 4A: Delay to Symptoms in NEC infants with large versus small bowel strictures.

Table 4B: Delay to diagnosis in NEC infants with large versus small bowel strictures.

1. Kosloske A (1994) Epidemiology of necrotizing enterocolitis. Acta Paediatrica83: 2-7.
2. Henry MC and Moss RL (2009) Necrotizing enterocolitis. Annu Rev Med 60: 111-124.
3. Neu J and Walker WA (2011) Necrotizing Enterocolitis. New England Journal of Medicine 364: 255-264.
4. Pourcyrous M, Korones SB, Yang W, Boulden TF, Bada HS (2005) C-reactive protein in the diagnosis, management, and prognosis of neonatal necrotizing enterocolitis. Pediatrics 116: 1064-1069.
5. Gaudin A, Farnoux C, Bonnard A, Alison M, Maury L, et al. (2013) Necrotizing enterocolitis (NEC) and the risk of intestinal stricture: the value of C-reactive protein. PLoS One 8: e76858.
6. Butter A, et al. (2002) The changing face of surgical indications for necrotizing enterocolitis. J Pediatr Surg37: 496-499.
7. Evrard J, Khamis J, Rausin L, Legat C, Bertrand JM, et al. (1991) A scoring system in predicting the risk of intestinal stricture in necrotizing enterocolitis. Eur J Pediatr 150: 757-760.
8. Isaacs D, North J, Lindsell D, Wilkinson AR (1987) Serum acute phase reactants in necrotizing enterocolitis. Acta Paediatr Scand76: 923-927.
9. Kenton AB, O'Donovan D, Cass DL, Helmrath MA, Smith EO, et al. (2005) Severe thrombocytopenia predicts outcome in neonates with necrotizing enterocolitis. J Perinatol 25: 14-20.
10. Lamireau T, Llanas B, Chateil JF, Sarlangue J, Cavert-Jouanolou MH, et al. (1996) Increasing frequency and diagnostic difficulties in intestinal stenosis after necrotizing enterocolitis. Arch Pediatr3: 9-15.
11. Ragazzi S, Pierro A, Peters M, Fasoli L, Eaton S (2003) Early full blood count and severity of disease in neonates with necrotizing enterocolitis. Pediatr Surg Int 19: 376-379.
12. Phad N, Trivedi A, Todd D, Lakkundi A (2014) Intestinal Strictures Post-Necrotising Enterocolitis: Clinical Profile and Risk Factors. Journal of Neonatal Surgery3: 44. 
13. Born M, Holgersen LO, Shahrivar F, Stanley-Brown E, Hilfer C (1985) Routine contrast enemas for diagnosing and managing strictures following non-operative treatment of necrotizing enterocolitis. J Pediatr Surg20: 461-463.
14. Kosloske AM, Burstein J, Bartow SA (1980) Intestinal obstruction due to colonic stricture following neonatal necrotizing enterocolitis. Ann Surg 192: 202-207.
15. Terrin G, Scipione A, Curtis MD (2014) Update in Pathogenesis and Prospective in Treatment of Necrotizing Enterocolitis. BioMed Research International1-9.
16. Schanler RJ (2016) Management of necrotizing enterocolitis in newborns.
17. Rees CM, Hall N, Eaton S, Pierro A (2005) Surgical strategies for necrotising enterocolitis: A survey of practice in the United Kingdom. Archives of Disease in Childhood - Fetal and Neonatal Edition90: F152-F155.
18. Pierro A (2005) The surgical management of necrotising enterocolitis. Early Hum Dev 81: 79-85.
19. Yoon JM, Park JY, Ko KO, Lim JW, Cheon EJ, et al. (2014) Fecal calprotectin concentration in neonatal necrotizing enterocolitis. Korean J Pediatr Korean Journal of Pediatrics57: 351-356.
20. Schwartz MZ, Joan Richardson C, Keith Hayden C, Swischuk LE, Tyson KRT (1980) Intestinal stenosis following successful medical management of necrotizing enterocolitis. J Pediatr Surg 15: 890-899.
21. Schwartz, M. Z., Keith Hayden C, Joan Richardson C, Tyson KRT, Lobe TE (1982) A prospective evaluation of intestinal stenosis following necrotizing enterocolitis. J Pediatr Surg17: 764-770.
22. Schimpl G, et al. (1994) Intestinal obstruction due to colonic stricture following neonatal necrotizing enterocolitisl strictures following successful treatment of necrotizing enterocolitis. Acta Paediatr Suppl 1994; 396: 80-83.
23. Janik JS, Ein SH, Mancer K (1981) Intestinal stricture after necrotizing enterocolitis. J Pediatr Surg 16: 438-443.
24. Griffin N, Grant L, Anderson S, Irving P, Sanderson J (2012) Small bowel MR enterography: Problem solving in Crohn’s disease. Insights Imaging3: 251-263.
25. Ha CY, Kumar N, Raptis CA, Narra VR, Ciorba MA (2011) Magnetic Resonance Enterography: Safe and Effective Imaging for Structuring Crohn’s Disease. Dig Dis Sci Digestive Diseases and Sciences56: 2906-2913.