Journal of Pediatric Surgery (2013) 48, 111117

www.elsevier.com/locate/jpedsurg

Probiotic prophylaxis after pullthrough for Hirschsprung disease to reduce incidence of enterocolitis: A prospective, randomized, double-blind, placebo-controlled, multicenter trial,,
Mohamed El-Sawaf b,1 , Sabina Siddiqui a,1 , Moustafa Mahmoud b , Robert Drongowski a , Daniel H. Teitelbaum a,
a b

Section of Pediatric Surgery , Mott Children's Hospital, University of Michigan, Ann Arbor, MI, USA Pediatric Surgery Unit, Department of Surgery, Tanta University, Tanta, Egypt

Received 25 September 2012; accepted 13 October 2012

Key words:
Hirschsprung disease; Enterocolitis; Probiotics; Aganglionosis

Abstract Objective: Hirschsprung-associated enterocolitis (HAEC) is one of the most troublesome problems encountered after a pullthrough. We hypothesized that prophylactic administration of probiotics after a pullthrough procedure would decrease the incidence of HAEC. Study Design: A prospective, double-blind, placebo-controlled, randomized trial was conducted at 2 children's hospitals. Infants undergoing pullthrough were randomized to probiotic or placebo for a period of 3 months post-pullthrough. Primary outcome was incidence of post-operative HAEC. Other outcomes included severity of HAEC by clinical grade, number of HAEC episodes and extent of aganglionosis. Pearson Chi Square analysis, as well as logistic regression, was used for statistical analysis. Results: Sixty-two patients were recruited (Sites: A=40; B=22). One was lost to follow up and one immediate post-op death was not included in final analysis. Probiotics were administered to 32 patients. Distribution of placebo/probiotics was equal between sites (P=0.858). Mean age at pullthrough was 6.58.1(SD) months. The incidence of HAEC was 28.3%. The incidence of HAEC was not statistically different between probiotic and placebo study groups. Conclusions: Incidence of HAEC was not reduced with prophylactic probiotics. Future studies are needed to better determine the etiology and possible ways of preventing this complex condition. 2013 Elsevier Inc. All rights reserved.

Study Funded and Placebo and Probiotics provided by: VSL Pharmaceuticals Inc., Sigma-Tau Pharmaceuticals, Inc, Townson, MD. The corporation had no influence on the reporting of our results. Registered with: ClinicalTrials.gov, identifier: NCT00630838. Presented at the American Pediatric Surgical Association, San Antonio, TX, May, 2012. Corresponding author. Section of Pediatric Surgery, University of Michigan Hospitals, Mott F3970, Box 0245, Ann Arbor, Michigan 48109. Tel.: +1 734 936 8464; fax: +1 734 936 9784. E-mail address: dttlbm@umich.edu (D.H. Teitelbaum). 1 Both the first and second authors contributed equally to the production of this manuscript.

0022-3468/$ see front matter 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpedsurg.2012.10.028

112 Hirschsprung-associated enterocolitis (HAEC) is one of the most troublesome problems encountered after a pullthrough. HAEC is a significant source of morbidity and potential mortality in this patient population, with a reported incidence of 17% to more than 40% during the postoperative course after pullthrough procedure [15]. The rate may reach as high as 55% in patients with total colonic aganglionosis Hirschsprung disease (HD) [6]. HAEC is the most common cause of death in children with HD [2,7,8]. Although the exact etiology of HAEC is unknown, an alteration in the gut microbiome or loss of intestinal epithelial barrier has been hypothesized to play a mechanistic role [5,9,10]. Adequate function of the intestinal mucosal barrier is also thought to play a major role in the prevention of HAEC, including the alteration of intestinal mucins [11]. HAEC may occur through a defect in this barrier, which may allow for the translocation of pathogens through the mucosa triggering the start of the inflammatory process. Probiotics are live microbes that, when administered in adequate amounts, confer a health benefit to the host. Probiotics potentially play a protective role in maintaining intestinal mucosal integrity through a number of different interactions, including alterations in mucosal cytokine expression, competing with intestinal pathogens for mucosal receptors, thereby increasing transepithelial resistance [12]. Probiotics have been suggested to be efficacious in the prevention and treatment of viral diarrhea [13], necrotizing enterocolitis [14] and pouchitis [15]. As well, probiotics can promote mucin production, which is deranged in HAEC [16]. Based on this previous knowledge on the beneficial effects of probiotics during pro-inflammatory conditions of the gastrointestinal tract, we hypothesized that prophylactic administration of probiotics after a pullthrough procedure would decrease the incidence of HAEC.

M. El-Sawaf et al. shortened life expectancy, or an inability to tolerate or take the oral probiotics. Further, as children on parenteral nutrition have been reported to translocate many probiotics into the blood stream, this was another exclusion in the postpullthrough period. The coordinators were able to approach 67 families. Of these, 62 families consented to participation and underwent randomization. Thirty-two patients were randomized to the probiotic group. One patient was dropped from the final study analysis due to early post-pullthrough death from unrelated causes (pneumonia and sepsis, one week postpullthrough) and one was dropped due to complete noncompliance (Fig. 1).

1.2. Study medication

VSL#3 (VSL Pharmaceuticals, Inc., Ft. Lauderdale, FL) is a probiotic preparation containing 90 billion viable lyophilized bacteria (per packet) composed of four strains of Lactobacillus, three strains of Bifidobacterium, and one strain of Streptococcus salivarius subsp. thermophilus. The complete preparation lactic acid bacteria and bifidobacteria, microcrystalline cellulose, stearic acid, magnesium stearate, and silicon dioxide in a vegetable capsule have been shown to be safe in children and neonates. The drug was delivered to the families in individual packets with maize starch as a filler. The probiotic was diluted in 2 to 4 oz of expressed human milk or formula. The placebo consisted of identical bags containing only maize starch. Study participants were administered either VSL#3 or maize starch orally twice a day. Dosing of the probiotic was based on body weight. Children b 5 kg received 90 billion bacteria (one 0.25 g sachet/dose) twice a day, and those over 5 kg received 180 billion bacteria (two sachets/dose) twice a day. Both compounds were indistinguishable in appearance.

1.3. Study design, masking and randomization

1. Methods
1.1. Patient recruitment
The study was performed with IRB approval (IRB HUM00000684) at the University of Michigan, Ann Arbor (defined as Site B) as well as by the Ethical Committee at Tanta University, Tanta, Egypt (defined as Site A).The study was registered with ClinicalTrials.gov, identifier: NCT00630838. The study coordinator at each site evaluated new hospital admissions for potential candidates during the recruitment period of 1/2006 to 12/2009. The diagnosis of HD was confirmed by suction rectal biopsy results prior to surgical pullthrough. Inclusion criteria consisted of all patients suffering from HD at an age of 48 months or younger. At the time of pullthrough parents/guardians were approached for study enrollment. Exclusion criteria were children N 48 months of age at time of pullthrough, hemodynamically unstable, active septicemia, major congenital anomalies with markedly

The study was a double blinded and randomization was conducted by the Investigational Drug Service in the Department of Pharmacy who was responsible of dispensing either probiotic or placebo. The dispensed medication was coded related to the patient ID and these codes were kept from the treating physicians and patient's parents. Block randomization with a randomly chosen block size of 2 or 4 within center was used. Randomization tables and envelopes were generated by the Department of Biostatistics at the University of Michigan and were distributed to each site. The patients were evaluated on a routine post-operative visit schedule of 1, 3, 6 and 12 months post pullthrough procedure as well as during unanticipated clinic, emergency room or hospital admissions. All data were available and collected for 60 patients. Compliance was measured by questioning the parents at post-operative follow-up, as well as having parents bring in empty sachets at the time of their clinic visit. Prospective data were collected, including feeding intolerance and incidence, frequency and severity

Probiotic prophylaxis after pull through for HD

113 reduce the incidence of Hirschsprung-associated enterocolitis from 40% in the placebo group to 10% in the treatment group. A model that tests the equality of two binomial probabilities was used, based on the estimated treatment effect and a power of 0.80. Specifically, this power analysis showed that we can achieve the desired power with sample sizes of 40 per group using a one-sided test with =0.05.

1.6. Statistical analysis

T-test was used for comparison between groups for the primary outcome measure. Multivariate regression was performed to analyze for contributing covariates predisposing to HAEC (e.g., gender, a diagnosis of Trisomy 21, length of aganglionosis, age at diagnosis, and age at pullthrough procedure). During the course of these experiments, an inspection of the data demonstrated an actual bias towards more HAEC in the probiotics group. Based on this, and the fact that a predicted additional 8 patients into each study would not influence the statistical outcome of the study, it was elected to terminate the study at a somewhat earlier point.

Fig. 1 Consort assignment of patients. The number of excluded patients included those who did not meet eligibility criteria and those whose parents or guardians refused to participate or secondarily dropped out of the study within the first 5 days after randomization.

of post-operative HAEC. Primary outcome measure was the occurrence of HAEC. HAEC was clinically diagnosed and documented by the treating physician. A common definition was based on previous diagnostic criteria at both centers [2]. Secondary outcome measures included number of HAEC episodes, the clinical severity of HAEC, and associated contributing factors predisposing to HAEC (e.g., gender, a diagnosis of Trisomy 21, length of aganglionosis, age at diagnosis, and age at pullthrough procedure).

1.7. Data safety monitoring

An independent data safety monitoring board was established which examined data halfway through the recruiting process, and did not identify any significant adverse events between study groups (e.g. death, sepsis, or severe diarrhea or feeding problems that were possibly associated with either treatment arm).

1.4. Evaluation and clinical grading of HAEC

To facilitate the diagnosis and objectively classify the severity of HAEC, the diagnosis was stratified into three grades (mild, moderate or severe) according to the previously described grading system by Elhalaby et al. [17]. (Table 1).

2. Results
2.1. Demographics
Sixty patients had completed charts and were used in the final analysis. (Site: A =40; B=20). Probiotic therapy was administered to 32 patients. Comparison of demographics between study sites is shown in Table 2. Distribution of the patients assigned to placebo versus probiotic groups was

1.5. Power analysis

Based on previous published series of HAEC, we modeled the study anticipating that probiotic prophylaxis would
Table 1 Grade I Clinical grading of HAEC.

Grade II

Grade III

Mild explosive diarrhea Mild to moderate abdominal distension No significant systemic manifestations (fever, anorexia, weight loss, tachycardia) Moderate explosive diarrhea Moderate to severe abdominal distention Associated with mild to moderate systemic manifestations Explosive diarrhea Marked abdominal distention Hypotension/shock or impending shock

114
Table 2 Demographic distribution by site of study. Site A (Tanta) (N=40) 29 (73%) 9 (23%) 4 (10%) 32.9 6.4 3 (8%) 5 (13%) 3 (7.5%) Site B (Ann Arbor) (N=20) 14 (70%) 2 (10%) 3 (15%) 7.4 2.6 10 (50%) 12 (60%) 9 (45%)

M. El-Sawaf et al.

Demographic Variable Gender (n, male) Prematurity (n) Diagnosis of Trisomy 21 (n) Age at diagnosis (monthsSD) Long segment aganglionosis a Incidence of Postoperative HAEC Recurrent Episodes of HAEC

p-value 0.839 0.238 0.570 0.009 b 0.001 b 0.001 b 0.001

Demographic Distribution by Randomization Group Demographic Variable Gender (n, male) Prematurity (n) Diagnosis of Trisomy 21 (n) Age at diagnosis (months) Long segment aganglionosis a Incidence of Postoperative HAEC Recurrent episodes of HAEC
a

Probiotic group (N=28) 25 (89%) 6 (21%) 1 (3.5%) 327.6 7 (25%) 10 (36%) 6 (21%)

Placebo group (N=32) 18 (56%) 5 (16%) 6 (18%) 284.9 6 (19%) 7 (22%) 6(19%)

p-value 0.235 0.929 0.068 0.491 0.967 0.592 0.796

Long segment was defined as proximal to the rectosigmoid region.

equal between sites (p=0.858). Gender distribution, number of premature patients, and the incidence of Trisomy 21 were similar between study centers. Seven children had an associated diagnosis of Trisomy 21. There were however, some significant differences between the two study sites. First, the age at diagnosis and at time of pullthrough procedure was significantly older in the Egypt group. Second, the length of aganglionosis was significantly longer in the Ann Arbor (B) group. Length of aganglionosis was defined as standard involving the rectosigmoid junction versus long segment involving colon proximal to the rectosigmoid junction, including total colonic Hirschsprung. Finally, the incidence of post-pullthrough HAEC was higher in the Ann Arbor group (Table 2). Despite these differences, it was important to note that the distribution of children with long vs. standard segment aganglionosis was similar between the placebo and probiotic groups (Table 2). The type of pullthrough operation, approach and existence of pre-pullthrough stomas were also noted to be evenly distributed (Table 3). The mean age of colostomy creation was 1.5 months old (range 2 days4 months). The most common indication for colostomy creation was active

enterocolitis (n=4), followed by poor general condition (n=2) and severe abdominal distention (n=2).

2.2. HAEC
The overall incidence of HAEC was 28.3%. Of the 28 patients randomized to the placebo group, 7 were diagnosed with post-operative HAEC (25%). Six were classified as clinical grade I, none as grade II and one as grade III (Table 3). Six of these seven patients went on to have recurrent episodes of HAEC (19%). Of the 32 patients randomized to probiotic treatment, 10 were diagnosed with post-operative HAEC (31%); and this was not significantly different from the findings in the placebo group. Three cases were classified as clinical grade I, six as grade II and one as grade III. Six of these ten patients went on to have recurrent episodes of HAEC (21%). The incidence of postoperative HAEC was noted to be more prevalent at Site B (Ann Arbor). This was thought to reflect the fact that long segment disease, a known risk factor for postoperative HAEC [18], was encountered more frequently in Site B. The overall incidence of post-operative

Table 3

Type of procedure performed by randomization group. Probiotic Group (n) Placebo Group (n) 5 14 7 3 3 1 p-value p=0.336 p=1.236

Presence of Pre-Pullthrough Colostomy Type of Pullthrough operation

Trans-anal pullthrough Open Soave Swenson's Lap Soave Other

4 16 8 3 3 0

Probiotic prophylaxis after pull through for HD

Table 4

115

Stratification of patients with and without HAEC based on randomization, severity of HAEC and length of aganglionosis. Placebo Site A Site B Y N Grade I Grade II Grade III Grade I Grade II Grade III HAEC Y HAEC N HAEC Y HAEC N 11 19 7 23 6 0 1 7 12 1 2 15 4 19 3 5 Probiotic 12 21 10 23 3 6 1 10 11 10 1 22 6 19 4 4 p value p=0.858 p=0.897 p=0.693

HAEC (N) HAEC Severity (First episode; N)

HAEC Severity (Second, third and fourth episodes; N)

p=0.648

Standard Length Aganglionosis (N) Long-Segment Aganglionosis (N)

p=0.925

HAEC was not significantly different between the patients receiving placebo or probiotic therapy (p=0.897; Table 3). Because of differences in the rate of HAEC between sites, as well as differences between the extent of aganglionosis, further stratification of the patients was then performed. We first stratified the study groups by severity of HAEC (See Table 3). Treatment with probiotics did not affect severity or clinical grade of HAEC experienced both for the first incidence of HAEC (p=0.693) or with a recurrence rate of HAEC (p=0.648).We next stratified HAEC by patients with standard length aganglionosis versus those with longsegment aganglionosis and found no differences in patients receiving placebo or probiotics (p=0.925). Multivariate logistic regression models for HAEC using treatment arms (p=0.592), gender (p=0.604), diagnosis of Trisomy 21 (0.072), age at diagnosis (b or N 1 week) (p= 0.198), and extent of aganglionosis (standard versus long segment) (p=0.021), showed a trend toward Trisomy 21 predisposing to HAEC and longer lengths of aganglionosis as a predisposing factor. However, regression analysis further demonstrated no significant influence on the effect of treatment with probiotics and its impact on the incidence of HAEC. HosmerLemeshow goodness-of-fit demonstrated no significant difference between the predicted and actual distribution, and thus the model was a good fit (Tables 4 and 5).

3. Discussion
The overall objective of this study was to test our hypothesis that prophylactic administration of probiotics after a pullthrough procedure would decrease the incidence of Hirschsprung-associated enterocolitis. Our study demonstrated that prophylactic treatment with probiotic therapy was ineffective in the prevention of HAEC. The exact etiology of HAEC remains unclear, despite advances in the diagnosis and management of HD. Some hypothesize that stasis due to a partial mechanical obstruction by the remaining aganglionic segment or spastic internal sphincter may be the inciting cause [1,3,19]. The loss of normal mucosal immune defense is another possibility, with several groups showing that there is a loss of normal mucosal immune defense, either due to the release of immunoglobulins in the aganglionic colon [7,20,21], altered microbiome [10] or a deficiency in the transfer of secretory IgA [7,21,22] as well as aberrant mucosal mucins [11,18,23] within the gastrointestinal mucosa resulting in a propensity towards infection. The prevailing model for the development of HAEC suggests a multifactorial etiology, due to a combination of normal mucosal defense loss followed by an invasion of pathogens into the intestinal wall with subsequent development of an

Table 5 Covariates

Relative risk of HAEC based on several influencing covariates. Relative risk 1.150 2.951 0.653 1.897 0.778 3.373 95% Confidence intervals 0.7021.883 1.1597.515 0.3811.117 0.4747.598 0.2952.053 0.84213.508 p-value 0.592 0.021 0.067 0.364 0.604 0.072

Probiotic Therapy Long segment aganglionosis Age at Pull Through (N 1 year old) Diagnosis of Trisomy 21 Female Gender Use of Antacid therapy

116 enterocolitic process. Many risk factors have been identified and help support the current model for the pathophysiology of HAEC. Children who have experienced delayed diagnosis of HD have been shown to have increased incidence of enterocolitis [4]. In addition children who develop HAEC are prone to recurrent episodes. The higher recurrence rate may be attributed to the persistence of the mucosal histopathological changes for long periods even after the performance of decompressive colostomy or even following successful pullthrough procedure [2]. Other risk factors include increased length of the aganglionic segment, female sex, Trisomy 21 and the presence of other associated congenital anomalies [2,4,9]. Of which long segment aganglionosis, and possibly age > 1 year at time of pullthrough and Trisomy 21 were predisposing factors for HAEC in the present study. The treatment of HAEC consists of rectal decompression along with oral antibiotics. In very ill patients intravenous antibiotics are used. The period of treatment varies from patient to patient; usually 1 to 2 weeks of treatment is required. It is of utmost importance that the treatment of enterocolitis is started early and pursued vigorously to prevent the condition from becoming chronic or spreading systemically. Refractory and repetitive enterocolitis is a common and serious problem in pediatric surgical patients, especially in patients with long-segment aganglionosis. Currently there is no single strategy identified for prophylaxis from this major complication. However, an increasing body of literature supports the use of probiotics to prevent or treat a number of pro-inflammatory conditions of the gastrointestinal tract. These include, pouchitis [24], inflammatory bowel disease [25] and the prevention of necrotizing enterocolitis [14]. Based on these previous studies, our hypothesis that probiotics would reduce the incidence of HAEC seemed quite plausible. While this was a prospective, randomized, controlled trial, there were some potential confounding factors that may have influenced the results. First, our study showed that despite our two centers being geographically quite distinct, our outcomes were not markedly different; and neither center found probiotics to be able to prevent the development of enterocolitis. It was interesting that the age at presentation was quite older in the Tanta, Egypt site, yet the overall rate of enterocolitis was lower than the Ann Arbor group. Further analysis showed that the Ann Arbor site had a higher incidence of long segment disease. As long segment disease has been associated with higher rates of HAEC [2], this may have influenced the comparative rates of HAEC seen between sites. Other practice differences existed between study sites that may have played a role. It was interesting that the Ann Arbor group was more likely to treat these children with antacid therapy while the Tanta group did not have any of its patients treated with either H2 blockers or proton pump inhibitors. The adult literature describes the use of antacid therapy as an independent risk factor for the development of

M. El-Sawaf et al. community acquired colitis and is posited to be due to an alternation in the microbiome of the gastrointestinal tract [26]. The use of these medications may have put the Ann Arbor cohort at further risk of developing enterocolitis and may have even played a role as a confounding factor influencing the outcome. As the exact mechanism of action of probiotic therapy is yet to be understood, it is difficult to draw conclusions as to why this modality of treatment was unable to diminish the incidence or severity of HAEC in patients post-pullthrough procedure. It is possible that other formulations of probiotics may have had a more positive effect, although VSL#3 contains a broad range of probiotic strains and one of the highest concentrations of probiotics commercially available. Further, because HAEC occurred more frequently in infants in the probiotic group, we do not anticipate any probiotic having a positive impact on reducing HAEC. We may benefit from repeating the study with a more stringent control over concurrent medical regimens, including suppressive antibiotics and anti-reflux therapy. In conclusion, the results of this study indicate that the use of probiotic bacterial therapy was ineffective in reducing either the incidence or severity of post-pullthrough HAEC.

References
[1] Caneiro P, Brereton R, Drake D, et al. Enterocolitis in Hirschsprung's disease. Pediatr Surg Int 1992;7:356-60. [2] Elhalaby E, Coran A, Blane C, et al. Enterocolitis associated with Hirschsprung's disease: a clinicalradiological characterization based on 168 patients. J Pediatr Surg 1995;30:76-83. [3] Hackam DJ, Filler RM, Pearl RH. Enterocolitis after the surgical treatment of Hirschsprung's disease: risk factors and financial impact.[see comment]. J Pediatr Surg 1998;33:830-3. [4] Teitelbaum D, Qualman S, Caniano D. Hirschsprung's disease. Identification of risk factors for enterocolitis. Ann Surg 1988;207:240-4. [5] Pastor A, Osman F, Teitelbaum D, et al. Development of a standardized definition for Hirschsprung's-associated enterocolitis: a Delphi analysis. J Pediatr Surg 2009;44:251-6. [6] Wildhaber BE, Teitelbaum DH, Coran AG. Total colonic Hirschsprung's disease: a 28-year experience. J Pediatr Surg 2005;40:203-6 [discussion 206-207]. [7] Fujimoto T. Natural history and pathophysiology of enterocolitis in the piebald lethal mouse model of Hirschsprung's disease. J Pediatr Surg 1988;23:237-42. [8] Elhalaby EA, Teitelbaum DH, Coran AG, et al. Enterocolitis associated with Hirschsprung's disease: a clinical histopathological correlative study. J Pediatr Surg 1995;30:1023-6 [discussion 1026-1027]. [9] Teitelbaum DH, Coran AG. Enterocolitis. [Review] [59 refs]. Semin Pediatr Surg 1998;7:162-9. [10] De Filippo C, Pini-Prato A, Mattioli G, et al. Genomics approach to the analysis of bacterial communities dynamics in Hirschsprung's disease-associated enterocolitis: a pilot study. Pediatr Surg Int 2010;26:465-71. [11] Mattar AF, Coran AG, Teitelbaum DH. MUC-2 mucin production in Hirschsprung's disease: possible association with enterocolitis development. J Pediatr Surg 2003;38:417-21 [discussion 417-421]. [12] Khan MW, Kale AA, Bere P, et al. Microbes, intestinal inflammation and probiotics. Expert Rev Gastroenterol Hepatol 2012; 6:81-94.

Probiotic prophylaxis after pull through for HD

[13] Salari P, Nikfar S, Abdollahi M. A meta-analysis and systematic review on the effect of probiotics in acute diarrhea. Inflamm Allergy Drug Targets 2012;11:3-14. [14] Deshpande G, Rao S, Patole S, et al. Updated meta-analysis of probiotics for preventing necrotizing enterocolitis in preterm neonates. Pediatrics 2010;125:921-30. [15] Holubar SD, Cima RR, Sandborn WJ, et al. Treatment and prevention of pouchitis after ileal pouch-anal anastomosis for chronic ulcerative colitis. Cochrane Database Syst Rev 2010:CD001176. [16] Mattar A, Drongowski R, Coran AG, et al. MUC-2 mucin production in Hirschsprung's disease: possible association with enterocolitis development. J Pediatr Surg 2003;38(3):417-21. [17] Bartizal KF, Salkowski C, Pleasants JR, et al. The effect of microbial flora, diet, and age on the tumoricidal activity of natural killer cells. J Leukoc Biol 1984;36:739-50. [18] Teitelbaum DH, Caniano DA, Qualman SJ. The pathophysiology of Hirschsprung's-associated enterocolitis: importance of histologic correlates. J Pediatr Surg 1989;24:1271-7. [19] Teitelbaum DH, Cilley RE, Sherman NJ, et al. A decade of experience with the primary pull-through for hirschsprung disease in the newborn period: a multicenter analysis of outcomes. Ann Surg 2000;232:372-80.

117
[20] Wilson-Storey D, Scobie W, Raeburn J. Defective white blood cell function in Hirschsprung's disease: a possible predisposing factor to enterocolitis. J R Coll Surg Edinb 1988;33:185-8. [21] Wilson-Storey D, Scobie W. Impaired gastrointestinal mucosal defense in Hirschsprung's disease: a clue to the pathogenesis of enterocolitis? J Pediatr Surg 1989;24:462-4. [22] Caniano DA, Teitelbaum DH, Qualman SJ, et al. The piebaldlethal murine strain: investigation of the cause of early death. J Pediatr Surg 1989;24:906-10. [23] Aslam A, Spicer R, Corfield A. Children with Hirschsprung's disease have an abnormal colonic mucus defensive barrier independent of the bowel innervation status. J Pediatr Surg 1997;32:1206-10. [24] Gionchetti P, Rizzello F, Helwig U, et al. Prophylaxis of pouchitis onset with probiotic therapy: a double-blind, placebo-controlled trial. Gastroenterology 2003;124:1202-9. [25] Meijer BJ, Dieleman LA. Probiotics in the treatment of human inflammatory bowel diseases: update 2011. J Clin Gastroenterol 2011;45(Suppl):S139-44. [26] Dial S, Delaney JA, Barkun AN, et al. Use of gastric acid-suppressive agents and the risk of community-acquired Clostridium difficileassociated disease. JAMA 2005;294:2989-95.