header link home



World Congress of

Pediatric Gastroenterology,

Hepatology and Nutrition



October 5 – 8, 2016


Montreal, QC



Scientific Abstracts


FISPGHAN Member Societies:





Thursday, October 6, 2016




POSTER SESSION I - 12:00 – 2:00 PM












                                APFED OUTSTANDING EGID ABSTRACT AWARD

                                NUTRITION PRIZE



Friday, October 7, 2016



                                GERARD ODELL PRIZE

                                GRAND WATKINS PRIZE

Concurrent Session III – 10:00 AM


Concurrent Session III – 10:00 AM


                                APFED OUTSTANDING EGID ABSTRACT AWARD

Concurrent Session III – 10:00 AM


                                NEUROGASTROENTEROLOGY AND MOTILITY PRIZE:  CLINICAL (Supported by a grant from LABORIE)

                                NEUROGASTROENTEROLOGY AND MOTILITY PRIZE:  BASIC (Supported by a grant from MMS)

Poster Session II - 12:00 – 2:00 PM




                GLOBAL HEALTH










                                PANCREAS PRIZE





Saturday, October 8, 2016



                                                WILLIAM F. BALISTRERI PRIZE

                                                ENDOSCOPY PRIZE



POSTER SESSION III - 12:00 – 2:00 PM



                GLOBAL HEALTH











                ENDOSCOPY / IMAGING




Thursday, October 6, 2016



10:00 AM





Valeria Di Giovanni1, Robert Bandsma1, Celine Bourdon1, Christian J. Versloot1, Ling Zhang1, Ling Zhang1, Wieger Voskuijl2, John Parkinson1, 1Hospital for Sick Children, Toronto, ON, Canada, 2College of Medicine, Blantyre, Malawi

Context:  Mortality rates in children with severe acute malnutrition (SAM) remain high despite standardized rehabilitation protocols. Two forms of SAM are distinguished: marasmus and kwashiorkor. Marasmus is characterized by severe wasting, whereas kwashiorkor presents with nutritional edema and is characterized by more profound metabolic disturbances, including hypoalbumenia and a fatty liver. However, it is unknown if there are differences in the metabolic profiles between children with marasmus and kwaskiorkor and whether these differences could indicate the need for distinct clinical treatment plans for each form of SAM.

Objective: We aimed to 1) identify metabolic pathways which change due to nutritional rehabilitation and 2) determine if children with marasmus demonstrate different metabolic profiles from children with kwashiorkor.

Design: We studied 40 children with SAM (18 marasmus and 22 kwashiorkor) aged between 6 to 60 months, who were treated at Queen Elizabeth Central Hospital in Blantyre, Malawi. Using the Biocrates p180 kit for targeted metabolomics, we obtained measurements for 149 metabolites in serum at admission and prior to discharge after nutritional rehabilitation. Metabolites include 32 amino acids and biogenic amines, 14 acylcarnitines, 15 sphingolipids, 87 glycerophospholipids, and others.

Results: At admission, 8 amino acids, including 4 essential ones, differed between marasmus and kwashiorkor; these were all lower in children with kwashiorkor. However, with nutritional recovery, 17/21 amino acids were significantly increased and only tryptophan continued to be lower in kwashiorkor. Nutritional recovery increased only 4/12 biogenic amines which are related to cell cycle progression and oxidative stress. Again, kwashiorkor tended to have lower values; both kynurenine and total dimethylarginine continued to be lower in kwashiorkor compared to marasmus after nutritional rehabilitation. Sphingolipids were not altered by nutritional recovery and also did not differ between groups. At admission, most acylcarnitines (9 out of 14) were lower in kwashiorkor patients compared to those with marasmus, and 3 continued to be lower after nutritional recovery. Acylcarnitines were of particular interest as they relate to beta oxidation, energy metabolism, fatty acid transport and mitochondrial damage.

Conclusions: Many but not all metabolites increased following nutritional recovery, pointing to a restoration in metabolic homeostasis. At admission, metabolites levels that differ between the two forms of SAM are systematically lower in children with kwashiorkor. In particular, lower levels of acycarnitines in kwashiorkor patients point to potentially impaired beta oxidation of fatty acids, which can be a source of energy during malnutrition. Our results suggest that specific metabolic disruptions may underlie the different clinical manifestation of marasmus and kwashiorkor and could be the basis for differential targeted treatments.



Sam Cheng1, Lieqi Tang1, Shi Jin1, Steven Winesett1, Henry Binder2, 1University of Florida, Gainesville, FL, USA, 2Yale University, New Haven, CT, USA

Introduction: Treatment of infectious diarrheas remains a challenge globally, particularly in infants, young children, and immune-compromised patients. Children with infectious diarrhea who become dehydrated are normally treated with oral or intravenous rehydration. Although rehydration can replace the loss of fluid, it neither stops ongoing intestinal secretion nor does it reduce underlying gut inflammation. Therefore, there has been continuous effort to search for new cost-effective ways to safely stop diarrhea. The extracellular calcium-sensing receptor (CaSR) is a unique Class C G protein-coupled receptor that uses nutrients such as calcium, polyamines and aromatic amino acid as its ligands. Recent studies indicate that CaSR is highly expressed in the gut, and when activated by selective nutrients or specific chemical agonists, exhibits unusual pro-absorptive, anti-secretory, and anti-inflammatory properties. We therefore hypothesized in the present study that activating CaSR in the gut reduces secretory and inflammatory diarrheas.

Methods: To test this hypothesis, three models of diarrhea were induced in 4-6 week-old Sprague-Dawley rats and/or C57BL/6 mice (CaSR wild-type and knockout), and the effects of CaSR agonists calcium, spermine, tryptophan, and R568 were examined. These included 1) cholera toxin model of secretory diarrhea; 2) citrobacter model of infectious diarrhea; and 3) dextran sodium sulfate (DSS) model of inflammatory diarrhea. To further prove the concept, antidiarrheal effect of calcium, primary ligand of CaSR, was also assessed on patients with viral, bacterial, and/or parasitic enterocolitis.

Results: Mice receiving cholera toxin gavage developed secretory diarrhea; the latter was inhibited by R568 i.p. in wild-type but not in CaSR null mice. Similarly, mice receiving C. rodentum gavage or DSS treatment developed inflammatory diarrhea, which was significantly more severe in knockout mice than the wild-type controls. Increasing dietary calcium reduced the severity of diarrhea in wild-type mice; such an effect was not seen in CaSR mice. In rats, activation of CaSR by dietary calcium significantly delayed the onset, reduced the severity, and accelerated the recovery of DSS-induced colitis; so did dietary spermine and dietary tryptophan. Finally, five patients with infectious enterocolitis who developed severe diarrhea and hypocalcemia were selected to receive calcium replacement therapy. As calcium therapy continued and hypocalcemia improved, stool output decreased; when serum calcium levels normalized, diarrhea stopped.

Conclusion: These results suggest that targeting intestinal CaSR with nutrients, alone or in combination, may represent a new cost-effective method to stop diarrhea and treat inflammation in children. Clearly, randomized controlled trials are warranted. 



Andrea Lo Vecchio1, Alfredo Guarino1, Jorge Amil Dias2, James A, Berkley3, Chris Boey4, Mitchell B. Cohen5, Sylvia, Cruchet6, Eduardo Salazar-Lindo7, Sandhu Bhupinder8, Philip Sherman9, Toshiaki Shimizu10, Ilaria  Liguoro1, 1University of Naples Federico II, Naples, Italy, Italy, 2Hospital de São João,, Porto, Portugal, 3KEMRI/Wellcome Trust Research Programme, Kilifi, Kenya, 4University of Malaya, Kuala Lumpur, Malaysia, 5University of Alabama at Birmingham, Birmingham, AL, USA, 6INTA, Universidad de Chile, Macul, Chile, 7Universidad Peruana Cayetano Heredia, Lima, Peru, 8Bristol Royal Hospital for Children, Bristol, UK, 9Hospital for Sick Children, University of Toronto, Toronto, ON, Canada, 10Juntendo University Graduate School of Medicine, Tokyo, Japan

Background:  Since 2006 two effective and safe vaccines against rotavirus (RV) infection (Rotarix™ and RotaTeq®) have been recommended by WHO worldwide. In 2012, FISPGHAN identified the spreading of RV immunization (RVI) as a top priority for the control of diarrheal illness in childhood.

Aims: FISPGHAN Working Group (WG) on acute diarrhea aimed at estimating the current RVI coverage and identifying the major barriers to local implementation in all countries of the world.

Methods: A survey was distributed to national experts in infectious diseases and vaccinations between March 2015 and April 2016. The survey provided information on the inclusion of RVI in the National Immunization Plan, presence of RVI programs, costs and perception of local barriers to implementation.

Results: Among the 76 countries contacted, 44 provided a survey eligible for analysis (response rate 58%). RVI is recommended in 23/44 countries (52.3%) participatingin the survey. Although five countries have recommended RVI since 2006, most (13/44, 29.5%) included RVI in National Immunization Schedule between 2012 and 2014. The costs of vaccination are covered by the government (38.6%), by the GAVI Alliance (9%) or public and private insurances (6.8%) in some countries. However, in most cases, those costs are charged to families (43.1%).

The limited perception of RV severity by families (50%) and elevated costs (45.4%) are the major barriers for large-scale implementation of RVI program. Surprisingly, only 6 countries (13.6%) reported the timing of first administration within 6 weeks as a major barrier.

Conclusion: After approximately 30 years since the introduction of RVI, the implementation of this major life-saving intervention is still unacceptably low and remains a major target for reaching the Millennium Developmental goal.. To sustain and implement RVI, FISPGHAN could promote education for families/caregivers and physicians focused on the risk of RV diarrhea and efficacy of immunization.



Benedikte Grenov1, Hanifa Namusoke2, Betty Lanyero2, Nicolette Nabukeera-Barungi3, Christian Ritz1, Christian Mølgaard1, Henrik Friis1, Kim F. Michaelsen1, 1University of Copenhagen, Copenhagen, Denmark, 2 Mulago National Referral Hospital, Kampala, Uganda, 3Makerere University, Kampala, Uganda

Globally, undernutrition is the cause of approximately 3.1 million child deaths annually. Diarrhea is a major cause of morbidity and mortality in undernourished children. Probiotics seem to reduce duration of diarrhea in children, but results have mainly been obtained in well-nourished children in high-income countries. We aimed at investigating the effect of probiotics on diarrhea in children with severe acute malnutrition (SAM) during in- and outpatient treatment in a low-income country.

A randomized, double-blind, placebo-controlled study was carried out in 400 children admitted with SAM. Children received one sachet daily with a mixture of BB-12® and LGG® (10 billion colony-forming units/sachet, 50:50) or placebo during hospitalization and subsequent outpatient treatment for 8 – 12 weeks. The primary outcome was number of days with diarrhea during inpatient treatment. Secondary outcomes included number of days with diarrhea during outpatient treatment, diarrhea incidence and severity, pneumonia incidence, duration and severity, days with fever or vomiting, weight gain and recovery. All outcomes were analyzed separately for in- and outpatient treatment. Diarrhea data was collected using a stool diary in which caregivers noted every time their child passed stool and categorized the stool consistency according to a 4-point photo scale.

There was no difference in number of days with diarrhea during inpatient treatment for the probiotic vs. the placebo group (adjusted difference +0.2 days, 95% CI -0.8 to 1.2 days, p=0.69). However, during outpatient treatment the number of days with diarrhea was reduced in the probiotic group compared to the placebo group by 26% (adjusted difference -2.2 days, 95% CI -3.5 to 0.3, p=0.025). There was no significant effect of probiotics on diarrhea incidence or severity, pneumonia outcomes, fever, vomiting, weight gain or recovery. Although not significant, some outcomes related to infections occurred less frequently in the probiotic group during outpatient treatment (diarrhea incidence odds ratio (OR) 0.7 (0.4 to 1.2), p=0.17, pneumonia incidence OR 0.5 (0.2 to 1.3), p=0.17, fever -0.5 days (-1.3 to 0.2 days), p=0.16). Mortality was 13%  (n= 26) in the probiotic and 10% (n=20) in the placebo group (p=0.24).

LGG® and BB-12® did not reduce diarrhea during inpatient treatment of children with SAM, but reduced the number of days with diarrhea during outpatient treatment. This result is in line with a study testing a mixture of pro- and prebiotics in children with SAM which reported no effect of the intervention during inpatient treatment but a trend towards reduced mortality during outpatient treatment. Probiotics may have a future role in outpatient treatment of children with SAM.



M. Isabel Ordiz, Mark J Manary, Nurmohammad Shaikh, Indi Trehan, Phillip I Tarr, Washington University in Saint Louis School of Medicine, St. Louis, MO, USA

Background and Objective: Environmental enteric dysfunction (EED) can be assessed by the lactulose:mannitol (L:M) test. Our objective was to determine if selected host fecal transcripts were correlated with EED, and whether transcripts and clinical characteristics could be used to predict EED in rural African children.

Methods: Demographic and sanitation characteristics, along with L:M testing and host fecal transcript analyses from 798 asymptomatic Malawian children aged 12-61 months were compared to linear growth over the subsequent 3 months. Fecal host mRNA analysis included quantification of expression of 18 transcripts associated with L:M. Permeability was categorized as normal (L:M ≤0.15), moderate (0.15 < L:M < 0.45) and severe (L:M ≥ 0.45), and random forest predictive models were created.

Results: L:M was inversely correlated with linear growth over the subsequent 3 months (r   -0.32, p<0.001) and severe EED was associated with stunting (p<0.0001). Age <24 months, weight-for-height Z-score <0, domesticated animals in the child’s sleep environment, lack of a pit latrine or clean water source, and a recent history of diarrhea were associated with severe EED. A random forest model using CD53, HLA-DRA, MUC12 and TNF was 84% sensitive for severe EED and 83% sensitive for no EED.

Conclusion: Selected host fecal transcripts can be used in a random forest model as a non-invasive biomarker for categories of EED in rural African children.



Thokozani E. Phiri, University of Waterloo, Waterloo, ON, Canada                        

Early childhood adversity that impacts the growth of mothers in their childhood may negatively influence the growth outcomes of their young offspring. Famine studies in Holland and China have shown the impact of epigenetics. A study in Malawi investigates this phenomenon in the context of three meteorological droughts (1981, 1987, and 1992 respectively) and women residing in rural and peri-urban Mangochi district. A hypothesis is tested that there was a difference in mean LAZ, mean WAZ, and mean birthweight between children born to drought exposed mothers and children born to non-exposed mothers.

Being a natural experiment, women were already pre-selected into groups of those exposed and not exposed to three different droughts (1981, 1987, and 1992 respectively) by virtue of their date of birth (DOB). This retrospective cohort study used children’s neonatal size measurements taken during a randomized controlled trial (iLiNS-DYAD-M). Their mean birthweight, mean LAZ, and mean WAZ were the outcomes (n=1403 includes 12 twins) while their mothers’ environmental experience of drought was the exposure (n=1391). Some of the covariates were child sex and the maternal variables of education, height, BMI, marital status, sole head of household, “at risk” during pregnancy, and primiparity at “at risk” ages.

LAZ, WAZ and birthweight were positively associated with mother’s exposure to the drought of 1987 even with covariate analysis (p< 0.01) except for the birthweight which no longer showed statistically significant results. The droughts of 1981 and 1992 as independent variables were not associated with the study outcomes at any level of significance (alpha 0.01; 0.05; 0.1) even when covariates were added to the models. The covariates that were associated with LAZ and WAZ in the 1987 drought model were maternal height, being both a mother and head of household, and being an older or younger “at-risk” mother who was primiparous. Only maternal height had a positive association (p<0.01). Older mothers (age  >35 yr) or younger mothers (age >18 yr) deemed “at-risk” during pregnancy were more likely to have children with a relatively lower LAZ and WAZ  than mothers who were not at risk as did mothers who were heads of household. 

The 1987 drought appears to be different to the other droughts in that there were some noteworthy associations. It has been historically noted as being moderate while the 1992 one was the worst of the three; however, major drought relief efforts were implemented and offset much of the impact. In terms of the puzzling positive association observed with the 1987 drought and LAZ or WAZ, epigenetics may have played a role as reported in studies on the Great China Famine and the Dutch Famine. In sum, the present study may provide additional evidence for the reversal of the negative intergenerational impact of early childhood adversity.


Thursday, October 6, 2016



12:00 – 2:00 PM


* Poster of Distinction





Ajay Sharma, Neha Chawla, SJOG Midland Hospital, Perth, Western Australia

Short title: Pediatric Eosinophilic Esophagitis.

Keywords: children, eosinophilic esophagitis, and intervention.

Background: Eosinophilic esophagitis (EoE) is a chronic allergy/immune-mediated disease with significant morbidity. It is characterized by dense eosinophilic infiltration of esophageal epithelium. EoE is increasingly recognized in children with high prevalence in patients with atopic disease.

Objective: To report our experience in children with EoE.

Methods: We conducted a retrospective cohort study using prospectively collected data in children (0-18 years) with an eosinophilic esophageal disorder between July 2014 and October 2015. Clinical characteristics including personal and family history for allergy, peripheral eosinophilia and IgE levels, endoscopic and histologic findings, treatment details, and outcomes were studied.

Results: A total of 31 children were diagnosed with EoE during the short study period (mean age: 7.6 years; 24 males and 7 females). Presenting symptoms varied with age; 12 had atopic condition. On endoscopy, 29/31 had abnormal appearing esophageal mucosa. Linear furrowing and mild erythema were the most common endoscopic findings. The treatment included topical steroids and/or dietary therapy.

Conclusion: Our experience indicates that the diagnosis of eosinophilic esophagitis in children is on the rise, presenting symptoms vary with age and has a common pathophysiological background with allergy as reported in previous studies. Identifying causative allergen and monitoring response to treatment remains problematic. Treatment options include topical steroids or avoidance of food allergen. Dietary therapy, though safe, lacks strict compliance.



Alka Kumari1, Govind K Makharia1, Anil K Verma2, Università Politecnica delle Marche, Ancona, Italy, Shyam Prakash1, Prasenjit Das1, Mona Pathak1, V Sreenivas1, Vineet Ahuja1, Lalit Kumar1, 1All India Institute of Medical Sciences, New Dehli, India, 2Università Politecnica delle Marche, Ancona, Italy

Background/Aim: There is a need for non-invasive biomarker of enteropathy both for diagnosis and monitoring of enteropathic diseases such as celiac disease. Assessment of plasma citrulline and plasma levels of intestinal fatty acid binding protein (I-FABP) have been proposed, but their reliability is yet not confirmed. For confirmation of reliability of these markers, we used a model [patients with hematological malignancies receiving myeloablative therapy for hematopoietic stem-cell transplantation (HSCT)] where changes occur cyclically in rapidly dividing cells with myeloablative therapy.

Patients and Methods:  Seventy adult patients with hematological malignancies receiving myeloablative therapy for HSCTs were recruited. Plasma samples were collected at different time-points i.e., before and after transplantation on HSCT days -7, -5, -1, 0, +7, +15 and +28. Levels of plasma citrulline (by HPLC), I-FABP (by commercially available ELISA) and total leukocytes counts were measured at each point of time.

Results and discussion:  Concentration of citrulline in plasma decreased significantly and consistently below the baseline at day+7 (p<0.001), and its level rose gradually at day +15 and day+28. The concentration of I-FABP fluctuated between day +7 and day+15 and then it rose at day +28. The decrease in level of plasma citrulline followed similar pattern as observed by total leucocytes count in peripheral blood.

Conclusions:  The data suggests that plasma concentration of citrulline follows pattern of cyclical changes in enterocyte mass as induced by myeloablative therapy. Therefore, an assessment of plasma level of citrulline may prove to be a reliable marker of enterocyte mass.



Bridget C. Godwin1, Hiroshi Nakagawa2, Kelly Whelan2, Ben Wilkins1, Alain J. Benitez1, Maureen DeMarshall2, Gary Falk2, Jonathan Spergel1, Amanda B. Muir1, 1The Children's Hospital of Philadelphia, Philadelphia, PA, USA, 2University of Pennsylvania, Philadelphia, PA, USA

Background and Aims: Eosinophilic esophagitis (EoE) is a chronic, antigen and immune-mediated esophageal disease characterized by symptoms such as dysphagia and food impaction. Histologic changes in active EoE include a peak eosinophil count of eosinophils per high-power field (eos/hpf) and basal cell hyperplasia (BCH) in esophageal mucosa. According to current diagnostic criteria, EoE transitions from active disease to inactive disease when the peak eosinophil count drops below 15 eos/hpf, typically following therapeutic intervention. We aimed to evaluate BCH in EoE patients with inactive disease in relation to eosinophil count as well as patient-reported symptoms within 30 days prior to biopsy.

Methods: Pediatric and adult histology from an IRB-approved patient repository of esophageal biopsies was reviewed by a pathologist specializing in gastroenterology. Specimens were scored for eosinophils and BCH. These data were then correlated with clinical data obtained at time of endoscopy through the same IRB-approved study, specifically presence of symptoms within 30 days of endoscopy. 304 pediatric and 137 adult specimens were scored from a total of 204 pediatric and 107 adult patients. Patient age ranged from 1-66 years, with 75% of patients being male. Of the specimens reviewed and included for analysis, 52 were from non-EoE patients, 221 from patients with inactive EoE, and 74 from patients with active EoE. 94 patients were excluded from analysis given unclear diagnosis or status at time of endoscopy.

Results: Scoring revealed the presence of BCH (defined as basal cells that reached >20% of epithelial height) in 5.8% (n=52) of non-EoE patients, 24.9% (n=221) of inactive EoE patients, and 97.3% (n=74) of active EoE patients. Average BCH score was significantly higher in patients with inactive EoE than in non-EoE (p=0.0024). Eosinophil count in inactive EoE was significantly higher with persistent BCH (n=55, average eos/hpf 6.9 = 0.62) than without (n=166, average eos/hpf  2.2 +/- 0.25) (p<0.0001). In the inactive EoE cohort, symptoms such as heartburn, dysphagia and regurgitation were more prevalent in patients with BCH (53.7%, n=67) than patients without BCH (30.4%, n=07) (p=0.0007).

Conclusions: BCH persists in patients with EoE that is clinically defined as inactive. Persistent BCH correlates with patient symptoms and eosinophil burden in inactive EoE patients. These findings suggest that BCH may be used to evaluate symptoms in patients with inactive EoE.  Although current clinical EoE guidelines define inactive EoE disease as <15 eos/hpf present on esophageal biopsies, our data suggest that a lower eosinophil count than 15 may be necessary to truly define inactive disease. Further studies regarding basal cell hyperplasia and its role in inflammation and barrier defect will further characterize the importance of re-defining inactive disease.



Christian Boggio Marzet, Tilli María Anabel, Basaldúa María Teresa, Pirovano Hospital, Caba, Buenos Aires, Argentina

Introduction: CMA is a continually growing disorder. Several studies suggest that babies delivered via vaginal canal acquire the mother's vaginal microbiota, which may help protect them and promote a healthy immune system’ however, babies born via Caesarean section acquire bacteria from the hospital environment that may increase the risk of food allergies and other problems.

Aim: To evaluate Caesarean section as a risk factor to develop different forms of CMA presentation.

Methods: All children from 2010-2014 who were referred to the Pediatric Gastorenterology Section at Pirovano Hospital with a diagnosis of CMA were included. Treatment was provided depending on the treating physician. Medical records were used to recruit by sex, age, birth weight, gestational age, clinical presentation of CMA and type of delivery.

Results: 238 patients were included. Mean age 0.60 months ± SD 0.97 (range 0-48-0.72). Females 50.8%. Mean gestational age: 38.38 weeks ± SD 1.72 (range 32-42). Mean birth weight: 3.149 gr ± SD 595.19 (range 2.936-4.355). Clinical presentation: rectal bleeding (RB) 44.5%, reflux (GER) 19.3%, immediate reactions (IMM) 14.3%, enteropathy (ENT) 11.8% and colic (COL) 10.1%. Type of delivery: Caesarean 56.3%, vaginal 43.7%. No statistical significant differences were found between type of delivery and clinical presentation (p=0.70), type of delivery and sex (p=0.28) and sex and clinical presentation (p=0.62) (Chi square Test). A positive risk correlation was found between IMM and Caesarean section (OR 1.45 95% CI, 0.69-3.04) and between GER and COL and Caesarean section (OR 1.26 95% CI, 0.65-2.43 and OR 1.33 95%CI 0.55-3.17 respectively).

Conclusions: Caesarean section was found to be a risk factor for developing not only IMM but also delaying the reactions (GER and COL) of CMA.



Daniela Migliarese Isaac1, Seema Rajani1, Maryna Yaskina2, Hien Q. Huynh1, Justine M. Turner1, 1University of Alberta, Edmonton, AB, Canada, 2Women and Children’s Health Research Institute, University of Alberta, Edmonton, AB, Canada,

Introduction: Celiac disease (CD) is a common autoimmune enteropathy to gluten, leading to intestinal inflammation, villous atrophy, and malabsorption. CD screening involves anti-tissue transglutaminase (atTG) IgA levels, followed by intestinal biopsy for confirmation. A gluten-free diet (GFD) is required to alleviate symptoms, normalize atTG, and heal the intestinal mucosa in CD patients. CD monitoring includes following atTG titers post diagnosis. Limited pediatric studies exist examining the trend of atTG normalization, with no studies examining predictors of atTG over time in children with CD. We aimed to evaluate time to normalization of atTG in children post CD diagnosis, and to assess for independent predictors impacting this time.

Methods: A retrospective chart review was completed to evaluate atTG normalization time in pediatric CD patients diagnosed from 2007 to 2014 in the Stollery Pediatric Celiac Clinic (Edmonton, Alberta, Canada). The following clinical predictors were assessed for impact on time to normalization: initial atTG, Marsh score at diagnosis, GFD compliance (GFDC), age at diagnosis, gender, ethnicity, medical comorbidities, and family history of CD. Kaplan-Meier survival analysis was completed to assess time to atTG normalization, followed by Cox hazard regression to assess for independent predictors of atTG normalization time.

Results: 487 of 616 patients reviewed met inclusion criteria. Mean age was 9.3 years at diagnosis, with 64% females. Patients were followed from 6 months to 6 years. 80.5% of patients normalized atTG levels within the study time period. Median time to normalization was 407 days for all patients (95% CI [361-453]), and 364 days for GFD compliant patients (95% CI, [335-393]). Type 1 diabetes mellitus (T1DM) patients took significantly longer to normalize at 1204 days (95% CI, [199-2209], p<0.001). Ethnicity was associated with longer time to normalization, with South Asians (SA) taking 809 days (95% CI, [262-1356], p=0.001); however, the validity was poor due to wide differences in censoring between ethnicities (Caucasian 18.2%, SA 35.2%). T1DM (HR 0.363 [0.238-0.553], p<0.001) and higher baseline atTG (HR 0.999 [0.999-1], p<0.001) were significant predictors of longer time to atTG normalization on Cox hazard regression. Conversely, GFDC was a significant predictor of earlier normalization (OR 13.91 [7.859-24.621], p<0.001). The remaining variables assessed were not significant.

Conclusions: There is wide variation of rate and time to atTG normalization in children with CD. GFDC and lower atTG at diagnosis are predictors of earlier normalization. Patients with T1DM are less likely to normalize atTG levels, and have longer time to normalization. There may be an association between SA ethnicity and longer duration to normalization, but the validity of this finding is poor due to wide differences in censoring. Overall, there is a need for closer attention and education for these higher-risk populations.



Devendra I Mehta, Daniel Sanchez, Nishant Patel, Jeffrey Bornstein, Karoly Horvath, Jessica Bonilla, Chirajyoti Deb, Orlando, FL, USA

Background: Recent studies support a role of small intestinal fungal overgrowth ( SIFO)  in adults with IBS disease symptoms, immunocompromised or not. Commensal candida density has been reported to be <100 cfu/mL. We aimed to find the clinical significance of Candida detected by duodenal brushings in children undergoing diagnostic endocopy, and assess histology, candida staining, and mucosal disaccharidases to look for evidence of pathology.

Methodology: Out of a total of  111/1900 studies  had  had Candida overgrowth ( >1000 cfu/mL cfus) had their mucosal histology, associated diagnostic findings, and disaccharidase activity assessed in comparison with those without Candida. A subset of duodenal biopsy samples were examined microscopically with hematoxylin and eosin (H&E) and special periodic acid shift with diastase (PASD) staining. Statistical analyses between SIFO and 2500 sequential disaccharidases were done using IBM SPSS.

Results: One hundred eleven patients (age: 0 to 20+ years; female 61) had fungal overgrowth with differential counts ranging from 1 to 300x10^3 CFU/mL. Most of the fungal species were Candida but one had Trichosporon asahii overgrowth. Combined fungal and bacterial overgrowth (> 10^4) was present in 57 patients. 12 (12/111) patients had abnormal histology of which 3 (3/12) had focal villous blunting without any other known cause. 2 of these 3 samples revealed increased intraepithelial lymphocytes. 3 samples exhibited chronic duodenitis and one had Crohn’s disease. An examination of 30 patients with a PASD stain revealed no evidence of candida hyphae. Of these 33/111 patients with fungal overgrowth had normal disaccharidase activity, compared with 1454/2489 entire population (p<0.000), 17/111 had generalized depression compared with  124/2489 (p<0.000), 35/111 had lactase deficiency compared with 1045/2849 population (p<0.05).

Conclusion: Duodenal brushing sample was found to be suitable to detect fungal over growth at a threshold level of 1000 CFU/mL. Fungal overgrowth was caused by different Candida species. A significant number of patients with fungal overgrowth were found to have generalized depressed activities of all the disaccharidases and many of them have lactase deficiencies separate from generalized depression of disaccharidases compared with population rates. However, direct staining using PASD in a subgroup of 30 did not show evidence of invasion/ hyphae. Further studies are warranted to evaluate the association of fungal overgrowth and disaccharidase deficiencies and IBS symptoms. 



Sarah Donnelly1, J. Antonio Quiros1, Christine Carter-Kent1, James Marcin2, Yunru Huang2, Shaoli Sun1, Johanna Palmadottir1, 1MUSC Children's Hospital, Charleston, SC, USA, 2University of California - Davis Children’s Hospital, Sacramento, CA, USA

Mast cells are immune cells that secrete vasoactive mediators (histamine, tryptase) and are usually associated with anaphylactic responses. Increased numbers of mast cells have been found in both adults and pediatric patients with chronic abdominal pain. Mast cell activity has been recently linked with enteric nerve proliferation due to neuronal growth factors but its role in functional abdominal pain is unclear. There is currently limited research on determining the characteristics of pediatric patients with abdominal pain and increased intestinal mast cell counts, risk factors and outcomes to treatment.

Objective: To determine if there are any generalizable characteristics for pediatric (<18y) patients experiencing recurrent abdominal pain with increased gastrointestinal mast cell counts in order to better determine disease risks and outcomes from therapy.

Methods: We performed a retrospective study of all children and adolescents with abdominal pain who had staining to assess mast cell (MC) burden (n=42). MC specific stains (CD117) were requested in cases of refractory abdominal pain, with a history of headaches, flushing or hives. Non-parametric data on these patients was then gathered by chart review. Characteristics such as age, race, weight, allergy history, and location of residency were collected. Also treatment data and symptom reporting was collected for all subjects. We then ran statistical analyses to determine if any factors seemed to be a determinant of the presence of increased mast cell counts in intestinal biopsies, their anatomic location, and the prevalence of allergic disorders and response to treatment.

Results: From our sample population, 83% (n=35) of patients met biopsy criteria for increased mast cell presence (>20 MC/hpf) in at least one biopsy sample. Of these patients, 90.5 percent were white and 9.5 percent were black. There was a female predominance within the group at 64 percent. A large proportion of these patients were underweight (36%) and this was statistically significant in patients with colonic involvement (p-value 0.02). Of those patients who received allergy testing, 60 percent had positive results. 71% of positive patients were treated with either an H1 or H2 blocker, mast cell stabilizer, or leukotriene inhibitor and 68% of these had symptom improvement. 

Conclusion: The presence of clinical symptoms of mast cell activation in children and adolescents with chronic abdominal pain appears to be a significant indication for performing mast cell staining on biopsy. While typical hematologic markers for mast cell activation do not seem to correlate to increased GI mast cell presence, allergy testing may be indicated in those with positive biopsy results. Existing treatment options appear to have a substantial effect on symptom improvement but standardization of treatment strategies could enhance treatment outcomes and this might be an area for new drug development.



JP Franciosi1, LN Werk1, LK Handy1, J Crutchfield2, L Ingraham2, MC Diaz3, RA Gomez-Suarez1, J Hossain3, T Wysocki4, 1Nemours Children’s Health System, Orlando, FL, USA, 2Lockheed Martin Corporation, Orlando, FL, USA, 3Nemours Children’s Health System, Wilmington, DE, USA, 4Nemours Children’s Health System, Jacksonville, FL, USA

Infant gastroesophageal reflux (GER) is a common physiologic process that is frequently not distinguished from true infant gastroesophageal reflux disease (GERD). Non-adherence to clinical guidelines leads to over-diagnosis of gastroesophageal reflux disease, misdiagnosis of other conditions, over prescribing unnecessary medications and unnecessary medical testing. We conducted a retrospective electronic medical record review across the Nemours Children’s Health System’s general pediatrics practices over a 6 month period between 10/01/2015-3/31/2016 comprising 120 providers, 6249 unique infants and 21,823 total visits among infants 6 months of age or less. Among 919 unique infants with 1372 encounters over the time period, primary care practitioners recorded the following ICD-10 diagnoses: regurgitation (6.9%, 94 diagnoses/1,372 total encounters); GERD (55.8%, 766 diagnoses/1372 total encounters), other vomiting (5.5%, 75 diagnoses/1,372 total encounters); and colic (7.4%, 102 diagnoses/1372 total encounters). Proton pump inhibitor (PPI) medications were prescribed in 5.7% (52/919 unique infants) and H2 receptor antagonist (H2RA) were prescribed in 0.5% (5/919 unique infants). Among 18 upper GI radiology tests and 42 pyloric ultrasounds, 94.4% and 50.0% were respectively ordered for a diagnosis of GERD only. Infants were frequently given a diagnosis of GERD over regurgitation or colic, and PPI medications were prescribed for infants inconsistent with guidelines. We propose these findings indicate a gap between actual decision making and the widely disseminated AAP Clinical Report and the NASPGHAN Consensus Guidelines. An intervention to ensure clinicians are educated around best practices and that the knowledge is retained over time is discussed. We will explore the impact of promoting knowledge retention on decision making related to diagnosis, prescription of medications and request for medical testing.



Joanne Masterson1, Kathryn A. Biette1, Juliet A. Hammer1, Eoin N. McNamee1, Kelley E. Capocelli2, James J. Lee3, Glenn T. Furuta2, 1University of Colorado Denver, Aurora, CO, USA, 2Children’s Hospital Colorado, Aurora, CO, USA  3Mayo Clinic, Scottsdale, AZ, USA

Background:  Dysphagia, stricture and food impactions are common symptomatic consequences for eosinophilic esophagitis (EoE) patients. However, to date limited understanding of the pathophysiologic mechanisms of remodeling in EoE exits. We hypothesized that a hypersensitivity reaction with prolonged and localized eosinophilic inflammation in mice would lead to esophageal remodeling similar to that observed in EoE patients.

Methods: Transgenic mice overexpressing the eosinophilopoietin IL-5, selectively in the esophageal epithelium (L2-IL5), were sensitized and challenged with oxazolone 9 times/29 days (chronic: L2-IL5-OXA-C) and compared to a previously published EoE model where mice are challenged 3 times/8 days (acute: L2-IL5-OXA-A). Treatment with Dexamethasone (Dex: 200ug/ms i.p.) 3 times/week was used as a therapeutic intervention. H&E stained esophagi were evaluated using a histological score (index of epithelial inflammation [eosinophil & chronic inflammatory infiltrate], epithelial remodeling [basal zone hyperplasia (BZH), dilated intracellular spaces, hyperkeratosis], internal esophageal layer fibrosis and intramuscular inflammation. Immunohistochemistry was performed evaluating fibrosis (Massons Trichrome) and proliferation/BZH (Ki67). Esophageal tissues were processed for Taqman mRNA analysis, and immune cell infiltrates were characterized by flow cytometry.

Results: Significant esophageal histopathological alterations were observed in the L2-IL5-OXA-C mice compared to L2-IL5-OXA-A counterparts, including eosinophilia (6 vs. 3.8, P 0.01) microabscesses (1 vs. 0.75, p=0.1) and basal zone hyperplasia (BZH) (2.8 vs. 2, p<0.001). Increases in dilated intercellular spaces (DIS) (3.2 vs. 0.5, p<0.001) and internal esophageal layer fibrosis (IELF) (5.6 vs. 1, p<0.001) were observed in L2-IL5-OXA-C mice compared to L2-IL5-OXA-A. The overall histological score was significantly higher in L2-OXA-C mice (22.6 vs. 11.8, p<0.001). L2-IL5-OXA-C mice had increased Trichrome staining and expression of fibrotic remodeling genes including COL1A1 (3-fold, p<0.05) and COL4A1 (7-fold, p<0.05), while barrier dysfunction was indicated by decreased CLDN1 expression. Flow cytometry confirmed a significant increase is esophageal eosinophilia in L2-IL5-OXA-C mice compared to L2-IL5-OXA-A (14.5x10^5 vs. 1.15x10^5, p<0.05). Treatment with dexamethasone significantly attenuated histopathologic indices (3.8 vs. 22.6, Dex vs. Saline, p<0.001), including eosinophilia, DIS and fibrosis. Of note, Ki67 immunohistochemistry indicated a significant decrease in BZH (15 vs. 66 cells/hpf, p<0.001). Eosinophilic chemokines CCL11 (80%) and CCL24 (93%) were decreased and tight junction molecules CLDN1 (5-fold, p<0.05) and CLDN7 (15-fold, p<0.001) were increased indicating restored barrier.

Conclusion: Collectively, these studies identify the L2-IL5OXA mouse as a novel preclinical model to study the epithelial barrier dysfunction, fibrosis and remodeling associated with chronic EoE.



John Pohl1, Raphael Firszt1, Barbara Chatfield, Fadi Asfour1, Catherine McDonald2, Amy Lowichik1, 1University of Utah, Salt Lake City, UT, 2Primary Children's Hospital, Salt Lake City, UT, USA

Background and Aims: Pediatric patients with cystic fibrosis (CF) and pediatric patients with eosinophilic esophagitis (EoE) can both present with feeding problems which can complicate clinical care. This study evaluated the association of eosinophilic esophagitis (EoE) and CF occurring in a single pediatric cystic fibrosis (CF) center.

Methods: Patients with EoE and CF from our center were identified from the electronic medical record and from the Cystic Fibrosis Foundation Patient Registry. Specific endoscopic and biopsy findings of EoE as well as clinical findings and treatment modalities were evaluated in these patients.

Results:  Five patients with CF and EoE were identified (1.8% of all patients) over a 13-year time period. The average age at the time of EoE diagnosis was 5.9 years (range 1 to 16 years). Younger patients presented with feeding problems; older patients presented with dysphagia. All patients were treated with proton pump inhibitor therapy, and four patients received oral fluticasone therapy. Two patients received a gastrostomy for elemental formula therapy. Most patients had a documented history of food allergies.

Conclusions: EoE can occur in pediatric patients with CF, and the prevalence potentially may be higher than the general population. Patients with CF and associated feeding problems may need further evaluation for EoE. 



Julia Fritz, Mariko Suchi, Diana Lerner, Medical College of Wisconsin, Milwaukee, WI, USA

Background: Herpes simplex virus (HSV) has long been recognized as a common cause of infectious esophagitis. While the majority of cases occur in immunocompromised hosts, cases have been reported in the immunocompetent individuals. Therefore, an immunodeficiency work-up is not indicated after diagnosis of HSV esophagitis in an otherwise healthy individual. To date, 9 cases have been reported of concomitant HSV esophagitis and eosinophilic esophagitis (EoE) suggesting a causal relationship between these two entities. Most reports of HSV esophagitis in immunocompetent children are prior to publication of diagnostic guidelines for EoE and do not report follow-up endoscopies. The aim of this study is to review all patients diagnosed with HSV esophagitis at our institution and identify co-morbid conditions before and after diagnosis.

Methods:  We conducted an IRB-approved retrospective review of HSV esophagitis diagnosed at a single large academic institution between 1/1982 and 3/2016. Pathology records were queried for the inclusion of “herpes” and/or “HSV”. Records including biopsies from the upper gastrointestinal tract were reviewed to identify patients diagnosed with HSV esophagitis by microscopic examination or viral culture. Medical records of these patients were retrieved for additional clinical information including presentation, treatment, and outcomes.

Results: Sixteen patients with HSV esophagitis were identified. Five cases were immunosuppressed at diagnosis. Of the 11 immunocompetent patients, 3 (27.3%) had an underlying genetic syndrome, 3 (27.3%) had no significant past medical history and no follow-up after symptom resolution. Of the 5 remaining patients (45.6 %), 1 had a history of EoE prior to HSV infection. Three patients experienced symptom recurrence after initial improvement following HSV treatment and on follow-up endoscopy within 6 months of infection were diagnosed with EoE. All of these patients had co-morbid atopic conditions. One patient had symptom recurrence and a follow-up EGD showed 15-16 eos/hpf in the single obtained esophageal biopsy while on proton-pump inhibitor therapy. At last follow-up in 2003 (3 years after HSV infection), this patient continued to be symptomatic.

Conclusions:  Our 34-year retrospective review supports prior studies that while immunocompromised patients are at highest risk of HSV esophagitis, this diagnosis should be considered in all patients with acute onset dysphagia/odynophagia. Among the immunocompetent patients, EoE was a frequent co-morbidity, with the majority diagnosed within 6 months of HSV infection. On biopsy specimens at the time of active HSV esophagitis, the number of eosinophils in the squamous mucosa was not high enough to suspect the diagnosis of EoE.  Therefore, clinical follow-up with endoscopy may be necessary for children with atopic conditions presenting with HSV esophagitis.  Further investigation is needed to clarify the relationship between EoE and HSV esophagitis.


Maria Fiorentin1, Alessio Fasano1, Kourtney P Nickerson1, Stefania Senger1, Marcelo Sztein2, 1Massachusetts General Hospital, Charlestown, MA, USA , 2University of Maryland, Baltimore, MD, USA                

Introduction: Typhoid fever is a common worldwide illness, transmitted by the ingestion of food or water contaminated with the feces of an infected person, which contain the bacterium Salmonella Typhi.

The World Health Organization estimates that 22 million cases of typhoid fever occur annually, resulting in ~200,000 deaths. In common with other enteropathogens, S. Typhi has developed means of breaching the mucosal epithelial barrier by usurping signaling mechanisms within host cells. At present, much remains to be uncovered concerning the host responses to S. Typhi infection. Available vaccines are moderately protective. Current therapeutic strategies include antibiotic treatment; however the frequency of antibiotic-resistant serovars is increasing worldwide. Additionally, in some individuals chronic S. Typhi colonization of the gall bladder culminates in gall bladder cancer, therefore highlighting the significant short- and long-term consequences of infection. Given the adverse consequences of S. Typhi infection there is a critical need to understand the mechanisms of the disease process and the host immune response for the development of efficacious and        cost-effective vaccines.

Materials and Methods: Terminal ileum biopsies were collected from donors for direct infection. Whole biopsy infections were conducted using micro-snapwell mounting of tissue followed by addition of S. enterica serovar Typhi 2a to the apical surface. Prior to infection, Typhi was grown in Luria Bertoni broth under static or shaking conditions. Upon infection, changes in trans-epithelial electrical resistance (TEER), cytokine release, gene expression, and cellular localization were assessed.

Results and Conclusions: Use of whole biopsy (WB) model identified specific contributions of the epithelium in response to Typhi infection as assessed by RNA-sequencing, qPCR, and cytokine secretion. Consideration for bacterial inoculum preparation revealed that Typhi grown under static conditions express the Vi antigen, as well as, SPI-1 and SPI-2 effector proteins. Therefore, static or shaking inoculums were applied on the ex-vivo tissue model.Differences in cellular association of bacteria were assessed using immunofluorescence and transmission electron microscopy. To identify how the gut mucosa responds to infection, apical and basolateral culture supernatants were collected for cytokine production analysis. Some of the key pro-inflammatory molecules, such as IL-8 and IL-6 appeared decreased in Salmonella-infected tissues, in line with our RNA-seq data showing Salmonella-induced immune suppression. Together, our data characterizes key aspects of terminal ileum response to Typhi infection addressing a critical gap in our current understanding of Typhoid fever pathogenesis.



Mark L. Tenzer, Michael H. Hart, Kristin M. Knight, Carilion Clinic, Roanoke, VA, USA

Background: Eosinophilic esophagitis (EoE), an allergic inflammatory disease, requires expensive, repeated, and invasive esophagogastroduodenoscopies (EGDs) and esophageal biopsies for diagnosis and treatment, resulting in substantial physical, psychological, and financial impact. Previous EoE biomarker research found targets that are weakly predictive when studied independently. Using multivariate machine learning algorithms to combine biomarker and clinical data (features) into one model, we trained a support vector machine (SVM) algorithm to develop a powerful and statistically supported diagnostic tool. A previous study utilizing large retrospective and simulated datasets evaluated which machine learning algorithms and biomarkers performed best, identifying the feature set for this prospective study with a predicted specificity of approximately 0.99.

Methods: Patients presenting to pediatric gastroenterology with known or suspected EoE and planned EGDs and biopsies were enrolled. Diagnoses and survey symptomology, standard-of-care complete blood count (CBC), allergen-specific immunoglobulin E (IgE), and pathological biopsy data were collected. Eotaxin-3 and eosinophil-derived neurotoxin (EDN) levels in patients’ serum were determined through commercially available ELISAs. SVM was trained with some patients to learn to diagnose EoE, as defined by biopsy, and tested with others to assess whether the model was correct. Accuracy, sensitivity, and specificity were calculated using leave-one-out cross-validation (LOOCV): with n patients, SVM was trained with n-1 patients and tested with one patient left out. This procedure was repeated for each patient, and the results were averaged. Then, different combinations of features were temporarily removed from the model, which was re-trained, and the feature set producing the maximum LOOCV accuracy was used for the final model.

Results: Thirty patients were enrolled; seven patients undergoing treatment or missing data were excluded, resulting in a dataset of 23 patients   (7 EoE, 16 control). The final model incorporated eotaxin-3 and EDN ELISAs, 18 survey questions, IgE averaged across all allergens, 15 CBC results, and 8 symptoms, with LOOCV accuracy, sensitivity, and specificity equal to 1: all patients were correctly diagnosed with only non-invasive data. Statistical significance was assessed with an approximate permutation test (n=10,000). For accuracy, sensitivity, and specificity respectively, the approximated pvalues were 0.0000, 0.0005, and 0.0011.

Conclusions: Algorithms utilizing non-invasively obtained data successfully diagnose EoE, potentially alleviating the substantial diagnostic burden on patients and cost of care. Additional enrollment and future studies would provide further validation and refinement of the model, as well as establish a method for non-invasive, quantitative monitoring of EoE patients.



Terri Giordano, Joe Piccione, Karen Zur, Asim Maqbool, Albert Kim, Matthew J. Ryan, The Children's Hospital of Philadelphia, Philadelphia, PA, USA 

Introduction: Although there is controversy surrounding the use of impedance testing affecting outcomes of children undergoing laryngotracheal reconstruction (LTR), we propose routine gastroesophageal reflux disease (GERD) screening allows for treatments that significantly improve outcomes in children undergoing LTR. 

Laryngoscopic findings of GE reflux include erythema, edema, nodularity, ulceration, granuloma and cobblestoning. Evidence suggests a connection between GERD and the development of subglottic stenosis, and studies in animal models has shown that gastric juices negatively effect mucosal healing after subglottic injury. We have developed a multidisciplinary aerodigestive team and any children undergoing LTR will have a preoperative diagnostic evaluation with esophagogastroduodenoscopy (EGD) and impedance testing. 

This study’s primary objective was to determine if the EGD and impedance testing results impact treatment decisions in patients undergoing preoperative LTR evaluation. 

Methods: An observational case series with retrospective electronic medical record review was conducted at The Children's Hospital of Philadelphia analyzing existing patients’ preoperative LTR evaluation data from January 2008 through June 2014. 

The primary endpoint was to determine the utility of EGD and esophageal impedance monitoring leading to changes in medical management such as adjustment of medication doses, avoidance of potential food allergens or surgical interventions prior to LTR.

Results: Seventy-four subjects were included in the study; 29 (39%) were female and 45 (61%) were male. The subjects had a median of 2.8 years. Sixty (81%) subjects had subglottic stenosis and 63 (85%) had a tracheostomy tube. Sixty-one (82%) subjects reported GERD symptoms and 7 (9%) reported none. Twenty-two (30%) had a previous fundoplication.

Of the 74 subjects, 64 (86%) on gross appearance had normal EGD results, 4 (5%) had esophagitis and 7 (9%) had gastritis. Fifty-nine patients (80%) had normal esophageal biopsies. Impedance testing revealed that 13 (18%) had an abnormal acid:non-acid ratio.

EGD and impedance study results lead to change in treatment in 24 (32%) subjects. Eighteen (75%) subjects had a new medication added. Of the 18 subjects that had a new medication added, 1 (5.6%) was prescribed an H2 blocker, 15 (83.3%) were prescribed proton pump inhibitors, 2 (11.1%) were prescribed erythromycin, 1 (5.6%) was prescribed bethanechol and 1 (5.6%) was prescribed Biaxin.  Two subjects had multiple medications added. Seven (29.2%) had their current GI medication increased. One (4.2%) subject had tube feedings changed to post-pyloric feeding. No subjects had their medications discontinued.

Conclusion: Abnormal EGD and impedance results lead to a treatment change in 32% of patients in this study population. Fundoplication may not be protective in all centers and no subjects with normal EGD and/or impedance results had medications withdrawn.



Maxim Itkin, David Picolli, Yoav Dori, Children's Hospital of Philadelphia, Philadelphia, PA, USA   

Objective:  To review the outcome of liver lymphatic and intestinal lymphatic embolization in patients with PLE, in whom leak of the intestinal or liver lymph was demonstrated.

Methods:  This is a retrospective review of medical records and imaging of 5 consecutive patients (M/F 3/2, mean age 14. 8 y) with PLE. Three patients presented after Fontan surgery and two patients post-thoracic duct (TD) ligation for chylothorax. Lymphatic imaging included heavy weighted T2 MRI and DCMRL, intranodal lymphangiogram (IL) and liver lymphangiogram. Abnormal lymphatics were percutaneously accessed and embolized with Lipiodol or n-BCA glue.

Results: In 3 patients after Fontan surgery, liver lymphangiography demonstrated leak of liver lymph into the duodenum. Embolization of liver lymphatics with lipiodol in the first 2/3 patients was performed with temporary increase in albumin and improvement of symptoms in one patient, however complicated by duodenal bleeding. Embolization of the liver lymphatics with n-BCA in 1 patient resulted in temporary normalization of the serum albumin level and resolution of symptoms without complications. In 2 patients post-TD ligation lymphatic leak into intestine was demonstrated on DCRML and IL. Embolization of the intestinal lymphatics in these patients resulted in normalization of albumin level and resolution of the symptoms with no complications.

Conclusions: In this study, we demonstrated development of abnormal liver or intestinal flow with leakage of the lymph in duodenum in patients with PLE post Fontan surgery and TD ligations. Embolization of these abnormal lymphatics with the goal of stopping the lymph loss in intestine is a promising new treatment for this disease.



Michelle Tobin, Susan Schuval, Stony Brook Children's Hospital, Stony Brook, NY, USA                     

Eosinophilic esophagitis (EoE) is an emerging disorder of children and adults. Its natural history is largely unknown. Although the prevalence of this disorder is increasing, the number of patients diagnosed in early childhood remains low, and there is a paucity of clinical and epidemiologic data regarding young children with this disease. We have identified a cohort of 27 children diagnosed with EoE aged 6 years and younger. The mean age of onset of gastrointestinal symptoms was 10.9 months, with 46% of children reporting “symptoms since birth.”  The most common symptoms included vomiting (63%), failure to thrive (41%), choking/gagging (33%), abdominal pain (30%) and feeding difficulty (22%). On average, these patients initially presented to the pediatric gastroenterologist at 25.9 months, at which time 67% were diagnosed with gastroesophageal reflux disease. Forty-seven percent had a known food allergy, and 37% had atopic dermatitis. Upper endoscopies (esophogastroduodenoscopy/EGD) and biopsies were performed on average at 38.5 months of age.  Biopsy results revealed a mean eosinophil count of 36 eos/hpf in the mid-esophagus, and 55 eos/hpf distally. The mean interval between onset of symptoms and EoE diagnosis was 25 months. Thus, even though young children with EoE have early onset of symptoms, there still remains a significant delay before EGD, definitive diagnosis, and most importantly, the initiation of treatment. Pediatric gastroenterologists should have a high index of suspicion for EoE in young atopic children presenting with nonspecific upper gastrointestinal symptoms and should consider performing an EGD more promptly to assess for this disease.



Nicole Heinz, Samuel Nurko, Anne C Mudde, Madeleine Stout, Rina Wu, Kitzia Colliard, Eitan Rubinstein, Edda Fiebiger, Center for Motility and Functional GI Disorders, Boston Children's Hospital, Boston, MA, USA  

Introduction: Eosinophilic esophagitis (EoE) is a chronic, inflammatory esophageal disease, characterized by eosinophilic inflammation and esophageal dysfunction. Upper gastrointestinal (GI) symptoms in EoE are age-dependent. Adults and adolescents typically experience dysphagia and esophageal food impaction (EFI), while the occurrence of EFI in younger children is less common and poorly investigated. There is limited information regarding the natural history of EFI in adolescents.

Objectives:1)To characterize a cohort of pediatric EoE patients who experienced EFI as the presenting symptom of the disease.

           2)To describe the management of EFI patients presenting to the emergency room.

Methods:  A retrospective chart review was performed of pediatric EoE patients diagnosed at 18 years or younger. Patients who presented with EFI leading to EoE diagnosis were compared to controls who did not present with dysphagia or a history of EFI, and the management of the EFIs was examined. Controls were age-matched to EFI patients at the time of EoE diagnosis. Patients with achalasia, esophageal atresia, or tracheo-esophageal fistula were excluded.

Results: 32 children with EoE that experienced EFI, and 26 age-matched controls were identified. In 81% of EFIs the impacted food was meat. Clinical characteristics are presented in Table 1, and EFI management is shown in Figure 1. Average age at EoE diagnosis in the EFI group was 13.9 ±3.0 years (6.7-17.8 years) and was 13.5 ±3.0 years (6.7 – 17.8 years) in the controls. 29/32 patients presenting with EFI were male vs. 13/26 controls (90.6% vs. 50%, p=0.001). Endoscopy results were obtained in 28/32 patients following the EFI: 42% underwent an endoscopy within 30 days post-EFI (0-252 days, mean  75.8 ± 82.9 days). No EFI patient had a stricture on endoscopy, 3 had esophageal narrowing, and 1 had a Schatzki ring.

32 patients presented with EFI. 7 passed spontaneously, and 25 presented to the ER for further management. In the ER, 7 received glucagon or nitroglycerin, 12 underwent upper endoscopy, and 6 passed without intervention. Three failed management with muscle relaxant and went on to have an upper endoscopy. 15 upper endoscopies were done in total, and 4/15 revealed no further EFI. 10/25 (40%) EFIs presenting to the ER resolved spontaneously.


•  Food impaction resolved spontaneously in 40% of the children visiting the ER. This finding could have implications for the clinical management of food impaction in an emergency setting.

•  Patients that present with EFI leading to an EoE diagnosis are likely to be male. Clinical characteristics of pediatric patients presenting with and without EFI are similar to previously published data in the adult population.



Paul Dobria1, Kiran Gorla2, Alan Schwartz1, Thirumazhisai S. Gunasekaran2, James Berman2, 1University of Illinois at Chicago, Chicago, IL, USA, 2Advocate Children's Hospital, Park Ridge, IL, USA

Background: Pediatric Eosinophilic Esophagitis (EoE) is a chronic, waxing and waning disease which can present with symptoms that include abdominal pain and dysphagia. A previous study demonstrated that pediatric EoE patients are at risk for adverse quality of life and long-term clinical outcomes. The Pediatric Eosinophilic Esophagitis Symptom Score v2.0 (PEESS)—a validated symptom metric—consists of items designed to measure the frequency and severity of patient- and proxy-identified EoE symptoms. Long-term outcomes of pediatric EoE have not yet been evaluated using this instrument, nor have differences between the long-term symptoms of patients with abdominal pain predominant EoE (EoE-AP) and dysphagia predominant EoE (EoE-D). Aims: Using the PEESS, 1) investigate changes in clinical outcomes of pediatric EoE over time; 2) compare the long-term outcomes of EoE-AP and EoE-D patients.

Methods: Of the 210 EoE patients contacted, 100 (48%) participated. Parent proxy-reported symptoms were collected using the PEESS. Item response theory was used to obtain measures of frequency and severity of symptoms for all 100 patients. Change in frequency and severity of symptoms over time was investigated via regression analysis. Differential item functioning (DIF) analysis was employed to compare differences in symptoms between EoE-AP and EoE-D patient groups.

Results: For the full cohort of 100 patients [mean age: 14.3 ± 5.1 years, mean time since diagnosis: 4.1 ± 2.3 years, 79% male], on average, the measure of symptom frequency decreased over time (b1   –0.145, p=  0.035). Severity of symptoms remained stable over time (b1   0.002,          p=  0.98). 14 patients reported no symptoms. Of the 100 patients, 45 were classified as having EoE-AP, 36 as EoE-D, and 19 as other-symptom predominant. Results of the DIF analysis show that, in terms of frequency of symptoms, EoE-AP patients were more likely to report stomach pain (t   –3.59, p<0.001) and nausea (t   –2.09, p=  0.041) than EoE-D patients, who were more likely to report difficulty swallowing (t   3.13,   p=  0.003) and needing to drink a lot to help swallow food (t   2.67, p=  0.010). On the severity scale, EoE-AP patients were more likely to report severe stomach aches (t –2.34, p=  0.025), while EoE-D patients reported greater severity in needing to drink a lot to help swallow food             (t   2.22, p=  0.036).

Conclusion: In this long-term follow-up study, not all EoE pediatric patients reported persistent symptoms. Across all EoE subgroups, frequency of symptoms subsided over time, while severity remained constant. Stomach pain and nausea were more prevalent in EoE-AP patients, and difficulty swallowing was more frequently reported by EoE-D patients. Stomach pain was more severe in EoE-AP patients, while the need to drink a lot to swallow was more severe in EoE-D patients. Further study is needed to better understand the impact of treatment regimens on these two clinical EoE disease phenotypes and their outcomes.



Paul Harris, Carolina Serrano, Camila Palma, Miguel A. Leon, Macarena Vera, Caroll Hernandez, Escuela de Medicina, Pontificia Universidad Católica de Chile, Santiago, Chile               

Background and aims: H. pylori infection in children causes increased T regulatory cell responses, which down-regulate T cell-mediated inflammation. The CagA virulence factor is the main immunodominant protein in H. pylori strains. Previous reports in mice and adults suggest a role of this oncoprotein in the suppression of DC maturation in a manner dependent of IL-10. In this study, we sought to evaluate the effect of the CagA protein from H. pylori on pediatric monocyte-derived DCs (MoDCs) and T cell response.

Methods: Mononuclear cells were isolated from peripheral blood of non-infected pediatric donors (n=10 for DC, n=5 for T cell studies) by gradient centrifugation. CD14+ and CD4+ naïve T cells were purified by MACS. CD14+cells were cultured for 5 days in presence of IL-4 and GM-CSF to generate pediatric MoDCs. Then we analyzed MoDC for maturation/activation markers (HLA-DR, CD86, CD83) by FACS and cytokine secretion after a 48h exposure with H. pylori strains 26695 (western isolate), 26695 ΔCagA and Hpk5 (eastern isolate) at MOI of 10. In addition we analyzed Treg cell generation (CD4CD25Foxp3+) and cytokine secretion in 3 days DC:T cell co-culture  previously stimulated with the same three strains of H. pylori for 2 hrs.

Results: H. pylori exposed MoDC, regardless of CagA status showed a suppressive phenotype characterized by low levels of expression of CD83, CD86 and HLA-DR. Moreover high levels of IL-23 and IL-10 but not IL-12 were observed in H. pylori exposed DC irrespective of CagA virulence factor. In co-culture experiments, generation of inducible Tregs (CD4CD25Foxp3+) occurred in in low percentages in response to H. pylori exposed DCs. Cytokine secretion profiles in T cells was characterized by the expression of high levels of IFN-γ and IL-10 but not IL-17 and IL-13 regardless of CagA status.

Conclusion:  H. pylori induces a suppressive phenotype in pediatric MoDC with secretion of high levels of IL-10 regardless of CagA status of the strains. Further T cell differentiation from pediatric T naïve cells is also not determined by the presence of CagA and is characterized by a mixed Th1 and Treg profile. 

This work was supported by Fondecyt grants #1130387 and #11140232.



Pooja Mehta1, Angela M. Haas1, Nancy Creskoff Maune1, Taryn Brennan1, Michelle L. Henry1, Calies Menard-Katcher1, Shikha Sundaram1, Lucien Harthoorn2, Faith Takurukura2, Zhaoxing Pan1, Glenn T Furuta1, Dan Atkins1, 1University of Colorado School of Medicine; Children's Hospital Colorado, Aurora, CO, USA, 2Nutricia Advanced Medical Nutrition, Utrecht, Netherlands

Background: Gastroesophageal reflux disease (GERD) and eosinophilic esophagitis (EoE) have been associated with malnutrition and feeding dysfunction. The aim of this study was to compare the frequency of nutritional deficiencies and feeding dysfunction in children with GERD and EoE.

Methods: A prospective study enrolling children aged 1-17 with GERD or EoE was performed. Children were excluded with a dual diagnosis of GERD and EoE or other comorbidities associated with nutrition and feeding. Assessments included height, weight, 3-day food diary, and serum levels of ferritin, 25-OH vitamin D, parathyroid hormone (PTH), and prealbumin. Feeding dysfunction was characterized by the Behavioral Pediatric Feeding Assessment Scale (BPFAS), a validated measure of feeding dysfunction.  It consists of measures of child behaviors as well as parental feelings and strategies used to cope with these behaviors. Comparative means and frequencies were analyzed using Student’s t-tests, Chi-square, or Fisher exact tests. Spearman’s rank correlation coefficient was used to determine associations with BPFAS scores. Hedges’ g was calculated to compare BPFAS scores with those of a historical group of healthy controls.

Results: Participants were divided into a nutrition cohort (GERD n=47, EoE n=65) and a feeding cohort (GERD n=44, EoE n=73) based on inclusion/exclusion criteria. The mean weight-for-length (WFL) z-score of children younger than 36.5 months of age with GERD was -1.11 and EoE was -1.12 (p=0.92). The BMI Z-score for children older than 36.5 months of age with GERD was 0.10 and EoE was 0.23 (p=0.16). Both GERD and EoE children had normal intake of calories, carbohydrates, proteins, fats, and iron but Vitamin D intake less than the daily recommended intake. Both GERD and EoE children had normal mean ferritin (30 vs. 34 ng/mL), prealbumin (21 vs. 20 mg/dL), PTH (40 vs. 38 pg/mL), and Vitamin D (30 vs. 31 ng/mL) with no significant differences between GERD and EoE groups (p= 0.45, 0.35, 0.68, and 0.69 respectively). Thirty-two participants with EoE (49%) were on dietary elimination therapy. No significant differences in WFL z-score, BMI z-score, ferritin, prealbumin, PTH and vitamin D (p=0.22, 0.13, 0.20, 0.69, 0.97, and 0.64 respectively) were found when comparing children on and off dietary elimination therapy. There was no significant difference in BPFAS scores between GERD and EoE children with total BPFAS frequency scores of 81.5 for GERD and 76 for EoE (p=0.14) and problem scores of 8.4 for GERD and 9.7 for EoE (p=0.88). Both frequency and problem scores were higher than those of historical healthy controls (Hedges’g of 0.99 and 1.2 respectively).

Conclusion: A highly selected group of children with esophagitis are without nutritional complications. There may be benefit to providing anticipatory guidance to minimize mealtime challenges, monitoring for improvement, and referral to a feeding therapist if necessary.



Pornthep Tanpowpong, Noparat Prachasitthisak, Suporn Treepongkaruna, Chatmanee Lertudomphonwanit, Sophida Boonsathorn, Napat Angkathunyakul, Pattana Sornmayura, Wasun Chantratita, Faculty of Medicine Ramathibodi Hospital, Mahidol University, Bangkok, Thailand     

Background:  Cytomegalovirus causing gastrointestinal disease (i.e., CMV-GI disease) is a complication in immunocompromised patients which can lead to significant morbidity and mortality but can also be treatable. The current gold standard diagnostic investigation requires histopathology of mucosal biopsy demonstrating intranuclear/intracytoplasmic inclusion body. However, endoscopy and mucosal biopsy are considered invasive. Recently, studies on noninvasive markers such as stool or plasma CMV polymerase chain reaction (PCR) have been proposed to aid in the diagnosis of CMV-GI disease. Few reports on stool CMV PCR in immunocompromised adults showed promising results, and diagnostic values of plasma CMV PCR in CMV-GI disease demonstrate diverse findings. We evaluated the values of stool CMV PCR in the diagnosis of CMV-GI disease.

Objective: To determine the sensitivity, specificity and accuracy of stool CMV PCR, using histopathology as a gold standard, in the diagnosis of CMV-GI disease among immunocompromised children.

Study design: Cross-sectional diagnostic study.

Methods: We enrolled immunocompromised individuals (e.g., post-transplantation, receiving chemotherapy or high-dose corticosteroids) aged        < 20 years presented with gastrointestinal symptoms (e.g., prolonged diarrhea >10 days, lower or upper GI bleeding) at a tertiary care and teaching hospital from January 2015 to March 2016. We excluded children with uncorrectable coagulopathy, suspected surgical condition, or who received ganciclovir for ≥7 days. Stool samples were analyzed for quantitative CMV PCR by Abbott Real-time amplification with the limit of detection at 20 copies per mL. All underwent esophagogastroduodenoscopy and colonoscopy with tissue biopsies to evaluate histopathology for CMV.

Results: We had performed stool CMV PCR in 31 patients but two could not undergo endoscopy. Therefore, 29 patients were analyzed. Most were female (55%) with a median age of 75 months (IQR: 33,152). Post-liver transplantation (34%) was the most common underlying condition. Prolonged diarrhea > 10 days (37%) and lower GI bleeding (34%) were two most common presenting symptoms. Two stool samples showed inhibitors and were precluded from the final analysis. Among 27 patients, we found CMV-GI disease in 7 patients (26%).  The sensitivity, specificity and accuracy of stool CMV PCR were 71, 85, 82% respectively. Moreover, plasma CMV PCR was performed in all study children, which likely due to our institution’s practice in immunocompromised patients. The sensitivity, specificity and accuracy of plasma CMV PCR were 100, 86, 90% respectively (by using the cutoff value of 150 copies per mL). We found a significant correlation between stool and plasma CMV PCR (p<.001).

Conclusion: Stool CMV PCR may be used as a non-invasive tool for aiding in the diagnosis of CMV-GI disease. Plasma CMV PCR demonstrates significant correlation with stool CMV PCR and represents high diagnostic values.



Amir F. Kagalwalla1, Joshua B. Wechsler1, Pratibha G. Hotwagner1, Sally Schwartz1, Melanie M. Makhija1, Anthony P. Olive2, Carla M. Davis2, Seth Marcus3, Maria Manuel-Rubio1, Katie Amsden1, Kristin Johnson1, Maureen Sulkowski1, Jessica Ross1, Marion Groetch4, Mary Ellen Riffle4, Hector Melin-Aldana1, Deborah Schady5,, Barry K. Wershil1, Margaret H. Collins6, Mirna Chehade4, 1Ann & Robert H. Lurie Children’s Hospital of Chicago and Northwestern University, Feinberg School of Medicine. John H. Stroger Hospital of Cook County, Chicago, IL, USA, 2Texas Children’s Hospital., Houston, TX, USA, 3GI Care for Kids., Atlanta, GA, USA, 4Icahn School of Medicine at Mount Sinai., NY, NY, USA, 5Texas Children’s Hospital., Houston, TX, USA, 6Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA

Background and aims: Eosinophilic esophagitis (EoE) is a chronic immune-mediated disorder of the esophagus that is triggered by food antigen. Empiric six food elimination diet (SFED) is effective in inducing remission in children and adults and is recommended diet therapy for treatment of EoE. Milk, wheat, egg and soy were the four most common food triggers identified in children treated with SFED. Our hypothesis is that four food elimination diet (4-FED) excluding milk, wheat, egg and soy should induce histological, endoscopic, and clinical remission in a majority of children with EoE.

Methods: Children meeting the consensus guidelines for diagnosis of EoE were enrolled in this prospective multicenter study and empirically eliminated milk, wheat, egg and soy from their diet and after 8 weeks underwent upper endoscopy with biopsies to establish histologic remission. Histologic remission was defined as esophageal peak eosinophil count of <15 eosinophils per high power field (eos/hpf). Secondary endpoints were identification of specific food triggers.

Results: Eighty children (68% male, 9 years, 84% white, 90% atopic) were treated with 4-FED and histologic remission was demonstrated in 51 (64%) children with decrease in esophageal eosinophil count from 61±34 eos/hpf to 5±4 eos/hpf (p<0.0001) after treatment. One or more symptoms resolved in 78% of responders. Exudates improved in 96% (p<0.0001), edema in 67%, (p<0.0001), and furrows in 57% (p<0.0001). Milk (85%), wheat (32%), egg (31%), and soy (17%) were triggers identified. A single food trigger was identified in 59%, two foods in 20% and three foods in 5%. 

Conclusion: Empiric 4-FED by inducing remission in a majority of children with EoE is an effective dietary modality to treat EoE in children. It compares favorably with SFED.



Rafail I. Kushak1, Timothy M. Buie1, Katherine F. Murray1, Stephen B. Cox2, Caleb D. Phillips2, Naomi L. Ward3, Harland S. Winter1, 1Massachusetts General Hospital, Boston, MA, USA, 2Research and Testing Laboratory, Lubbock, TX, USA, 3University of Wyoming, Laramie, WY, USA

Autism is a neurodevelopmental disorder; however, many children with autism have gastrointestinal (GI) problems including constipation, diarrhea and abdominal pain that might affect behavior. GI problems have been associated with carbohydrate malabsorption (decreased lactase activity) or dysbiosis, alteration of the indigenous microbiota. Prior analysis of stool, colon, ileal and duodenal bacteria has revealed microbial dysbiosis, overgrowth of microorganisms, or depleted microbial diversity in autistic individuals relative to non-autistic populations.

Aims: To correlate disaccharidase activity (DA) and the duodenal mucosal microbiome of children with and without autism to determine if altering substrate availability in children with carbohydrate malabsorption affects the microbiome. 

Methods:  Activity of lactase, sucrase, maltase, and palatinase were evaluated in biopsies from 21 autistic subjects and 19 neurotypical subjects (controls) undergoing diagnostic upper endoscopy. Enzymes activity was evaluated using enzymatic assays. DNA was isolated from the duodenal biopsies, 16S rRNA was amplified via PCR; pyrosequencing was followed by computational analysis.

Results: There was no difference in DA between the two groups. In autistic group 17 out of 21 subjects and in controls 18 out of 19 subjects were lactase deficient. In samples from autistic subjects, the relative abundance of genus Bacteroides, Faecalibacterium, and Clostridium showed a statistically significant positive correlation with lactase activity. The duodenal microbiome in neurotypical children was different than in children with autism. None of the three groups of bacteria were observed to be correlated with disaccharidase activity in control samples, which instead showed strong positive correlations between lactase activity and the relative abundance of the genera Porphyromonas, Barnesiella, Gemella, and Leptotrichia. Correlation between activity of intestinal disaccharidases and abundance of microbiota was also found at the species level.

Conclusion: There are differences at the genus and species level in the duodenal microbiota in children with autism that could be influenced by maldigestion of lactose or nutritional differences in food consumption.



Romina Mehaudy, Marina Orsi, Claudio Parisi, Natalia Petriz, Hospital Italiano, Buenos Aires, Argentina    

Introduction:  The increase in food allergies in children is a worldwide concern. Studies on prevalence, incidence and the natural history of cow's milk allergy (CMA) are difficult to compare because of the deficiencies and inconsistencies in their designs. In the different series published, an incidence of 1% decreases to 0.3% when open challenges (thegold standard)   are used for diagnosis. That is why it is important to determine incidence by considering all immunological mechanisms involved.

Objective: To estimate the incidence of CMA in a hospital in Buenos Aires from June 1, 2015 until January 31, 2016.

Material and Methods: A prospective study was conducted in a population of an affiliated community hospital in Buenos Aires. Data were collected using electronic medical records. CMA incidence was calculated taking into account the proportion of children with CMA detected on the annual number of births.

The diagnosis was established according to clinical practice guidelines (DRACMA).

All patients with suspected CMA underwent skin prick test, patch test and open challenges to confirm the diagnosis.

Monthly monitoring of patients was performed, recruiting those who had suspected CMA.

Results: 768 infants were included from June 1, 2015 to January 31, 2016.

Of the 768 infants analyzed in our cohort, 51% were male, and 60% were born by Caesarean section.

13 patients with clinical suspicion of CMA showed an incidence of 1.7 % (95% CI, 0.9 to 3), when open challenges were performed; 8 cases were confirmed, showing a cumulative incidence of CMA of 1 % ( IC 95 from 0.4 to 2% CI).

Conclusion:  Conducting open challenges can confirm the presumptive diagnosis of CMA.

The systematic application of this test in patients with suspected non IgE -mediated CMA can reduce overdiagnosis and improve patients’quality of life.

Although these preliminary results are similar to those seen in other countries, this is the first study done in a Latin American Country.



Vincent Mukkada1, Nicholas Shaheen2, Gary Falk3, Christian Eichinger4, Helen Schofield4, Denise King4, Lora Todorova5, 1University of Cincinnati College of Medicine, Cincinnati, OH, USA, 2University of North Carolina School of Medicine, Chapel Hill, North Carolina, USA, 3University of Pennsylvania Perelman School of Medicine, Philadelphia, PA, USA, 4PharmaGenesis London, London, UK, 5Shire, Zug, Switzerland

Background and Aims: Eosinophilic esophagitis (EoE) is a chronic immune-mediated disease characterized by esophageal inflammation and dysfunction. To ensure a timely diagnosis of EoE in children, it is important for clinicians to be aware of common symptoms at presentation and potential differences across patient groups. This systematic review (SR) was designed to evaluate the literature on symptom patterns and progression of EoE in children.

Methods: Electronic databases (MEDLINE, Embase, Cochrane) and recent congresses were searched in February 2016 for studies describing the natural history of EoE in children.

Results: Of 1145 articles identified, 14 met the inclusion criteria. Data were available for 1830 patients who presented at centres in the US, Europe or Australia in 1982¨C2012. The mean age at diagnosis was 6 years. However, the average duration of symptoms prior to diagnosis was 1.2¨C3.5 yrs. The most commonly reported symptoms at presentation were dysphagia (5¨C61% of patients), emesis (17¨C60%), abdominal pain (16¨C55%) and food impaction (7¨C22%). Evidence from three studies suggests that symptoms may differ by ethnicity, age and sex. Failure to thrive (44¨C52% vs. 9¨C22%, both p<0.05), emesis (63% vs. 37%, p=0.005) and emesis/acid reflux (90% vs. 52%, p<0.01) were more common in African American (AA) than in Caucasian (Ca) children, while abdominal pain (27¨C29% vs. 40¨C57%, both p<0.05) and food impaction (10% vs. 26%, p=0.003) were less common. Eczema was also more frequent in AA children than in Ca children (57¨C52% vs. 9¨C20%, both p<0.05), while only one of two studies found asthma more common in AA compared with Ca children (51% vs. 32%, p=0.05). Additionally, symptom onset in AA children occurred at a younger age than in Ca children (3.7¨C5 vs. 9.1¨C10 years, p<0.01). Food impaction, dysphagia and heartburn were associated with age  ≥8 yrs, while emesis, growth failure and food refusal were associated with age <8 yrs (all p<0.05). Food impaction was more common in males (25 vs. 11%, p=0.016) whereas abdominal pain was more common in females (58% vs. 35%, p<0.001). After a mean follow-up of 7¨C15 yrs, 34¨C49% of adult patients still had difficulty swallowing. Resolution of symptoms was also rare, with the findings from one large study revealing that only 11/565 (2%) participants experienced complete resolution after a mean follow-up of 5.2 yrs. A second study found that untreated children experienced worsening of symptoms over a mean follow-up of 6 years.

Conclusions: Symptoms of EoE differ by ethnicity and age. Emesis and failure to thrive may be more common in AA than Ca children, and age at presentation may be lower in AA children. Inflammatory manifestations, such as emesis and growth failure, may be more common in younger children, while fibrotic manifestations, such as dysphagia and food impaction, may be more common in older children. 



Sally Lawrence1, Aoife Casey2, David C. Wilson2, Paul Henderson2, 1British Columbia  Children's Hospital, Vancouver, BC, Canada, 2Royal Hospital for Sick Children, Edinburgh, Scotland, UK

The clinical utility of fecal calprotectin in routine pediatric practice: a regional retrospective cohort study.

Background:Fecal calprotectin (FC) is elevated in children with inflammatory bowel disease (IBD). There is little information available regarding the use of FC in other pediatric gastrointestinal (GI) conditions.

Objectives: To determine the patient outcome of all clinical cases where FC was requested in routine pediatric practice by any pediatric specialist in a regional cohort of children. To evaluate the utility of a negative FC result when used in routine clinical pediatric practice.

Method:  Using a retrospective study design, all FC samples taken over a 6-year period between 2005 and 2010 in children <17 years old were obtained. All case notes were reviewed. Patients were followed up for a minimum of 5 years. FC was measured using the PhiCal Test (Calpro). Exclusion criteria were (1) known chronic GI disease; (2) age <1yr and ≥17yrs; (3) insufficient FC sample; (4) second or subsequent FC in the same diagnostic cycle; (5) FC taken following endoscopy.

Results: 728 patients were included (62% male, median age 8.2 (IQR 4.7-11.8) years with median follow-up 7.8 (range 5.3-11.2) years). 7% (49/728) were diagnosed with IBD, 18% (131/728) had non-IBD GI inflammation, 75% (548/728) had no GI inflammation (see table for diagnostic categories). Median FC was 1110ug/g (IQR 550-1630) for patients with IBD, 110ug/g (IQR 30-360) for non-IBD GI inflammation, and 30ug/g (IQR 20-70) for no GI inflammation. 6 of 49 (12%) IBD patients had FC <200; 3 with perianal disease (minimal luminal). 1.4% (10/728) of patients with initial FC <200ug/g developed IBD in a subsequent diagnostic cycle during a follow-up period of 14-124 months. The repeat FC (n=9) at the time of IBD diagnosis was >200ug/g in 70% (5/7); the other 2 patients had perianal disease with minimal luminal inflammation. 3.8% (28/728) of patients without a diagnosis of IBD had a FC >500ug/g; none developed IBD during follow-up but 75% (21/28) were diagnosed with non-IBD GI inflammation.

Conclusions: In routine pediatric clinical practice via a large regional cohort, FC <200ug/g rules out GI inflammation in 86% of symptomatic cases, which should help decrease endoscopy rates in this group. A minority (37%) of cases of GI inflammation are IBD, but the likelihood of IBD rapidly rises with higher FC values as FC results are higher in IBD than other causes of GI inflammation. Our study shows that an FC of >500 identifies mainly GI inflammation (non-IBD and incident IBD), but also a few cases without any evidence of GI inflammation even when followed up for a minimum of 5 years. Serial FC estimation is therefore important for patients with ongoing symptoms due to the possibility of either IBD or non-IBD GI inflammation in evolution.



Sanjay Kumar1, Dan Atkins2, Todd D. Nebesio1, Glenn T. Furuta2, Sandeep K. Gupta1, 1Riley Hospital for Children - Indiana University Health, Indianapolis, IN, USA, 2Children’s Hospital Colorado- University of Colorado School of Medicine, Aurora, CO, USA

Background: Eosinophilic esophagitis (EoE) is a chronic, immune-mediated condition of the esophagus. Topical corticosteroids (TCS), mainly fluticasone propionate and budesonide, are used for treatment of EoE. Recent literature has raised concerns about the possible association between chronic TCS use and risk of adrenal insufficiency (AI), a potentially life-threatening condition.

Aim: To assess practice patterns for AI screening of EoE patients on long term TCS.

Methods: A prospective study was performed to assess the practice patterns of gastroenterologists using a self-administered online survey. An email was sent to physicians (n=6) in various eosinophilic diseases consortia including, Consortium of Eosinophilic GI Diseases Research (CEGIR), The International Gastrointestinal Eosinophil ResearcherS (TIGERS), physician members of American Partnership For Eosinophilic Diseases (APFED) and Pediatric GI bulletin board listserv (n=2567), with a link to an online survey. The study was IRB approved, voluntary and no remuneration was given. Study participants were divided into 2 groups based on whether they are members of different EoE consortia  (C Consortia group, NC Non-consortia group). One survey was sent to an individual regardless of if they were a member of more than one consortium; i.e. ,only unique surveys were analyzed. Chi-square and Student’s  t-test were used for statistical comparison.

Results: One hundred fifty-one respondents started the survey of which 140 completed it. Twenty-one of 140 were from C group and 119 were from NC group. Overall, 14/140 (10%) respondents screen for AI. C group was statistically more likely to screen for AI as compared to NC group [9/21(43%) vs. 5/119 (4%); P 0.0001]. The majority of respondents [10/14 (71%)] use a morning cortisol level as the initial screening test. 6/14 (43%) respondents reported abnormal initial screening in >25% of patients. Referral to endocrinologist was made if initial screen is abnormal by 43% of respondents, whereas 28% repeat the screening test and 21% perform low dose ACTH stimulation test. Overall, 9/14 (64%) respondents reported one or more confirmed cases of AI.  Of those who do not currently screen for AI, 25/126 (20%) plan to screen for AI in the next 6 to 12 months, mainly due to recent literature.

Conclusions: C group members are more likely to screen for AI, possibly due to their expertise. While there is consistency in initial screening, there is considerable variability in follow-up testing. Even though an unexpectedly high number of respondents reported confirmed cases of AI, this is based on a small sample size of a selective group. However, 90% of respondents currently do not screen for AI, implying that AI may be an under-recognized condition in this patient population. Since 20% of respondents who currently do not screen report plans to change their practice pattern in next year, there is an urgent need for additional studies and formal AI evaluation guidelines for EoE patients on TCS.



Sarah Monks1, Peter Woodward2, Aparna Rao2, Shauna Schroeder2, Scott Schraff2, Phoenix, Lindsay Stevens2, Dana Williams2, 1University of Arizona - College of Medicine Phoenix, Phoenix, AZ, USA, 2Phoenix Children's Hospital, Phoenix, AZ, USA

Background: Dysphagia with aspiration (DA) is a common symptom in patients presenting for multidisciplinary gastroenterology, pulmonology, and ENT diagnostics at Phoenix Children Hospital’s Aerodigestive Clinic (ADC). DA is associated with chronic respiratory or gastrointestinal symptoms, chronic thickener use, constipation, and enteral tube dependency. MBS is routinely used to evaluate dysphagia severity and guide thickener treatment of DA patients and for re-evaluation after feeding therapy or other medical and surgical intervention. Previous ADC patients with deep interarytenoid notch (DIN) given injection laryngoplasty had improved symptoms despite post-intervention MBS scores worsening, and vice versa. We initiated a Quality Improvement (QI) project to determine if MBS severity is reflective of clinical symptoms of DA, aimed at changing dysphagia monitoring protocols in DA patients.

Methods: A clinical questionnaire of DA symptoms was developed and administered over 3 months to patients aged 1-3 years who had an MBS within 6 months of their initial ADC visit, standard of care for DA. 17 symptoms (12 gastrointestinal symptoms and 5 pulmonary symptoms) were given a numerical score 0-4 based on parent recall of frequency. MBS was scored 1-10 based on the thickness of liquid recommended for aspiration prevention (1 being thin and 10 being pudding consistency). Individual symptoms and symptom sets (total questionnaire score, GI symptom score, pulmonary score) were compared to MBS scores using linear regression model.

Results: 30 families of patients aged 1-3 were surveyed with median MBS score of 6 and range from 0 to 8. 18 patients had an MBS score above 6. 23 patients had more than one MBS evaluation performed before their initial presentation to the ADC, ranging from 2 to 4 previous evaluations. Median questionnaire score was 18 (range of 4 to 53). All 30 patients presented with at least 1 GI symptom and 29 patients presented with at least 1 pulmonary symptom. 8 patients had DIN. All analysis showed NO significant correlation between individual symptom or symptom sets and MBS score, with the highest R2 value for individual symptoms being 0.05.

Conclusions: Among ADC patients with symptomatic DA, MBS severity score did not correlate with severity and specificity of symptoms, questioning the use of MBS as a repetitive tool for diagnosing severity of persistent DA and for assessing response to laryngeal cleft surgical interventions and thickener wean therapy. In our patient population the use of repetitive MBS is challenged by these findings. Our team is developing a combined clinical and radiologic tool for long-term management aimed at minimizing radiation exposure while promoting best clinical outcomes.



Shiuh-Bin Fang1, Ke-Chuan Wang1, Chih-Hung Huang2, Ya-Chun Lee2, Tung-Cheng Chang1, Kee-Thai Kiu1, Yan-Hao Su1, Ming-Te Huang1, Hsu-Wei Fang2, 1Shuang Ho Hospital, Taipei Medical University, New Taipei City, Taiwan, 2National Taipei University of Technology, Taipei, Taiwan

Two currently licensed rotavirus vaccines, Rotarix and RotaTeq, are efficacious and safe in protecting children from rotavirus gastroenteritis. However, innate immune responses of human intestinal epithelia to rotavirus vaccination remain little known. Furthermore, live attenuated Salmonella strains have been developed as carriers of recombinant attenuated Salmonella vaccines (RASV) expressing heterologous antigens. This study aimed to investigate innate immune responses to rotavirus vaccines and the intracellular non-replicating S. Typhimurium ΔspeG mutant comprising plasmid encoding rotavirus VP4 and VP6 as a RASV in 5-d-old Caco-2 cells, in vitro M cells by coculturing Caco-2 cells and Raji B lymphocytes, and ex vivo human colonic mucosa using surgical explants established in polarized in vitro organ culture (pIVOC).
Non-polarized 4-d-old Caco-2 cells, polarized 20-d-old Caco-2 cells, and in vitro M cells were treated with Rotarix, RotaTeq, Salmonella Typhimurium ΔspeG, RASV, or none for 18 h, and human colonic mucosa ex vivo were treated with Rotarix, RotaTeq, or none for 5 h. The mRNA and protein expression levels of interleukins (IL-8 and/or IL-4, IL-6, IL-15) and hBD-2 were quantified using qRT-PCR and ELISA, respectively; the expression levels of untreated controls and treated groups were compared using the Student’s t-test in cell lines and Mann-Whitney test in pIVOC.  In 5-d non-polarized Caco-2 cells, Rotarix and RotaTeq significantly increased the protein and mRNA expression levels of IL-4, IL-6, IL-8, and IL-15 (all p<0.05, Table 1), but not hBD-2, whereas ΔspeG and RASV significantly increased both expression levels of IL-4, IL-6, IL-8 and hBD-2 (all p<0.05), but not IL-15. In 20-d polarized Caco-2 cells and in vitro M cells, the mRNA expression levels of IL-4, IL-6 and IL-8 were significantly upregulated in the Rotarix-treated Caco-2 cells, whereas those of IL-6 and IL-8 were significantly upregulated in the Rotarix-treated in vitro M cells. In addition, hBD-2 expression was non-significantly downregulated in the Rotarix-treated Caco-2 cells and in vitro M cells compared to untreated controls. RASV significantly induced higher apical secretion of hBD-2 from in vitro M cells than its vector ΔspeG. In pIVOC, Rotarix, but not RotaTeq, significantly suppressed hBD-2 mRNA expression of human colonic mucosa compared to untreated controls (1.00 [0.06-2.64] vs. 0.09 [0.05-0.19] fold change, p=0.041). No significant regulation of IL-6, IL-8, and IL-15 was found.

In conclusion, rotavirus vaccines can induce human innate responses in non-polarized Caco-2 cells, polarized Caco-2 cells, and in vitro M cells, particularly expression of IL-6 and IL-8. We successfully demonstrated hBD-2 suppression in ex vivo human colonic mucosa and a similar tendency in in vitro M cells after rotavirus vaccination for the first time. Our establishment of in vitro M cells and pIVOC can be developed as a platform to study novel oral vaccines .



 4d nonpolarized Caco2 cells (*p<0.05 compared to NT control, n=3)

ELISA (mean±SE)

NT control






IL4 (pg/mL)

49.6 ± 0.5

75.6 ± 3.7*

66.6 ± 1.1*

74.1 ± 1.4*

74.2 ± 2.4*

56.0 ± 0.01*

IL6 (pg/mL)

10.5 ± 5.1

47.0 ± 4.1*

55.9 ± 5.1*

248.8 ± 4.5

246.7 ± 5.4*

70.6 ± 4.4*

IL8 (pg/mL)

17.5 ± 4.1

61.2 ± 1.4*

58.3 ± 2.9*

181.8 ± 2.8

188.8 ± 14.4*

252.8 ± 12.6*

IL15 (pg/mL)

31.6 ± 5.6

53.1 ± 0.8*

54.7 ± 2.8*

50.4 ± 1.9*

46.8 ± 2.3

47.0 ± 1.6

hBD2 (pg/mL)

25.9 ± 3.6

28.6 ± 1.2

22.9 ± 2.8

781.6 ± 71.

807.9 ± 11.2*

455.4 ± 45.8*


NT control






IL4 (fold change)

1.0 ± 0.1

4.9 ± 0.7*

4.3 ± 0.3*

4.6 ± 0.3*

5.4 ± 0.4*

2.6 ± 0.2*

IL6 (fold change)

1.0 ± 0.2

3.2 ± 0.4*

2.6 ± 0.2*

6.3 ± 1.0*

5.9 ± 0.6*

2.8 ± 0.3*

IL8 (fold change)

1.0 ± 0.1

1.9 ± 0.3*

3.3 ± 0.4*

10.6 ± 1.6*

10.4 ± 1.6*

3.8 ± 0.4*

IL15 (fold change)

1.0 ± 0.1

1.3 ± 0.2*

1.4 ± 0.1*

1.2 ± 0.3

1.0 ± 0.1

1.4 ± 0.2*

hBD2 (fold change)

1.0 ± 0.1

1.1 ± 0.1

0.8 ± 0.1

10.8 ± 1.6*

10.6 ± 1.8*

4.5 ± 0.5*

20d polarized Caco2 cells (*p<0.05 compared to NT control, n=3)


NT control






IL8 (pg/mL) basolatera

0 ± 0

1.7 ± 1.4

0 ± 0

12.8 ± 2.9*

20.5 ± 3.9*

13.3 ± 3.2*

hBD2 (pg/mL) apical

0 ± 0

0 ± 0

3.3 ± 2.9

17.2 ± 14.9

11.4 ± 9.8

13. 1± 7.9

IL8 (pg/mL) cell

0 ± 0

0 ± 0

0 ± 0

31.0 ± 5.2*

30.0 ± 1.6*

0 ± 0

hBD2 (pg/mL) cell

0 ± 0

6.4 ± 3.3

2.1 ± 1.8

0 ± 0

0 ± 0

79.8 ± 9.5*


NT control






IL4 (fold change)

1.0 ± 0.6

5.3 ± 1.2*

0.7 ± 0.2

IL6 (fold change)

1.0 ± 0.2

6.7 ± 1.5*

3.9 ± 0.5*

IL8 (fold change)

1.0 ± 0.2

2.6 ± 0.1*

6.0 ± 0.5*

IL15 (fold change)

1.0 ± 0.1

1.5 ± 0.2

1.4 ± 0.2

hBD2 (fold change)

1.0 ± 0.2

0.6 ± 0.1

249.9 ± 66.4*

20d in vitro M cells (*p<0.05 compared to NT control; †compared to ΔspeG, n=3)


NT control






IL8 (pg/mL) basolatera

0 ± 0

16.2 ± 3.4*

11.4 ± 3.3*

23.2 ± 1.8*

32.3 ± 7.0*

49.8 ± 4.9*

hBD2 (pg/mL) apical

0 ± 0

0 ±0

0 ± 0


10.3 ± 8.9†

61.7 ± 18.7*

IL8 (pg/mL) cell

11.0 ± 4.0

15.0 ± 3.8

21.0 ± 4.0

13.0 ± 4.9

14.0 ± 6.1

37.0 ± 6.7*

hBD2 (pg/mL) cell

14.4 ± 10.2

0 ± 0

2. 3 ± 2.0

5.3 ± 4.6

7.9 ± 6.9

53. 7 ± 5.9*


NT control






IL4 (fold change)

0.8 ± 0.4

3.4 ± 2.0

0.8 ± 0.4

IL6 (fold change)

1.0 ± 0.1

5.5 ± 1.0*

3.9 ± 0.2*

IL8 (fold change)

0.8 ± 0.1

2.9 ± 0.3*

6.9 ± 2.6

IL15 (fold change)

1.1 ± 0.0

2.2 ± 0.5

1.3 ± 0.2

hBD2 (fold change)

1.2 ± 0.4

0.5 ± 0.1

202.5 ± 67.5*

Human colonic mucosa in pIVOC (*p<0.05 compared to NT control, n=6)


NT control



hBD2 (median, range)

1.00 (0.062.64)

0.09 (0.050.19)

0.56 (0.162.9)

IL6 (median, range)

1.00 (0.076.19)

0.75 (0.024.99)

0.66 (0.032.46)

IL8 (median, range)

1.00 (0.054.13)

0.64 (0.042.57)

0.90 (0.071.97)

IL15 (median, range)

1.00 (0.172.64)

0.09 (0.050.50)

0.56 (0.162.90)



Stefania Senger1, Kourtney P Nickerson1, Laura Ingano1, Marcelo Sztein1, Maria Fiorentino2, Alessio Fasano1, 1Massachusetts General Hospital, Charlestown, Massachusetts, USA  2University of Maryland, Baltimore, MD, USA

Introduction: Salmonella enterica serovar Typhi is the causative agent of Typhoid fever, from which an estimated 22 million cases occur annually resulting in 200,000 deaths. In some areas of the globe, the incidence of Typhoid fever is as high as 500 cases out of every 100,000 children. At present, little is understood about host response to Typhi infection. As such, no long-term preventive vaccine therapy is available. Current therapeutics include antibiotic treatment, however antibiotic resistant serovars are increasing worldwide. Furthermore, chronic Typhi colonization of the gall bladder is sufficient to cause gall bladder cancer, therefore demonstrating significant health risks upon both long- and short-term infection. To design alternative therapeutic strategies, there is immediate need to understand Typhi infection and pathogenesis.

Materials and Methods. Terminal ileum biopsies were collected from donors for generation of organoid culture. Culture conditions were optimized to maintain terminal ileum epithelial stem cells; cells were then seeded onto transwell inserts for generation of monolayers. Monolayers were infected with Salmonella enterica serovar Typhi 2a to the apical surface. Upon infection, changes in trans-epithelial electrical resistance (TEER), cytokine release, gene expression, and cellular localization were assessed.

Results and Conclusions: Terminal ileum derived organoids give rise to a diversity of epithelial cells, including goblet, paneth and M cells, which are grown as a monolayer in vitro. Use of the epithelial monolayer model identified specific contributions of the epithelium in response to Typhi infection as assessed by qPCR, immunofluorescence and cytokine secretion. Differences in cellular association of bacteria were assessed using IF and TEM. To identify how epithelium responds to infection, apical and basolateral culture supernatants were collected for ELISA analysis.  ELISA analysis demonstrated differences in IL-8, IL-1β and IL-12p70 cytokine production, with significant levels of basolateral cytokine secretion.  Finally, infection decreased TEER at 120m post infection. Together, our data characterizes key aspects of terminal ileum response to Typhi infection addressing a critical gap in our current understanding of Typhoid fever pathogenesis.



Steve B Min1, Cade M. Nylund2, Thomas P. Baker1, Mazer Ally 1, Brian Reinhardt1, Yen-Ju Chen1, Luz Nazareno1, Fouad J. Moawad11Walter Reed National Military Medical Center, Bethesda, MD, USA, 2Uniformed Services University of Health Sciences, Bethesda, MD, USA 

Background: The diagnosis and management of eosinophilic esophagitis (EoE) often requires multiple endoscopies. Serum biomarkers can be elevated in EoE patients, but their clinical utility is not well established.

Goals: To evaluate serum biomarkers in EoE subjects compared to controls and followed longitudinally in response to treatment.

Methods: We conducted a prospective cohort study of children and adults undergoing esophagogastroduodenoscopy (EGD) for suspected EoE.  After completing a course of proton pump inhibitor therapy, esophageal mucosal biopsies were obtained, as well as serum analysis of absolute eosinophil count (AEC), eotaxin-3, eosinophil derived neurotoxin (EDN), eosinophil cationic protein (ECP) and interleukin-5 (IL-5). Subjects with normal endoscopic and histologic findings constituted controls. Those meeting criteria for EoE underwent repeat EGD and biomarker measurements following treatment with topical steroids for 8 weeks.

Results: AEC (263.50 cu/mm vs. 102 cu/mm, p< 0.001), ECP (26.98 ng/mL vs. 5.20 ng/mL, p< 0.001) and EDN (31.70 ng/mL vs. 14.18 ng/mL, p=  0.004) levels were significantly elevated in EoE subjects compared to controls and correlated with esophageal eosinophilia. The level of AEC (odds ratio [OR], 1.79; 95% confidence interval [CI], 1.28-2.64) and ECP (OR, 1.61; 95 % CI, 1.23-2.36) were associated with a diagnosis of EoE. Only AEC significantly predicted esophageal eosinophilia following topical steroid therapy in EoE subjects (p=  0.006).

Conclusion:  AEC, ECP, and EDN were higher in EoE subjects compared to controls and correlated with degree of esophageal eosinophilia.   Furthermore, AEC predicted post-treatment eosinophilia, suggesting a potential role in monitoring EoE disease activity.



Suchitra K Hourigan, Nicole Clemency, Wendy S.W. Wong, Elisabeth Klein, Marina Provenzano, Ramaswamy Iyer, John E Niederhuber, Pediatric Specialists of Virginia, VA, USA

Background: Fecal occult blood cards (FOBC) that can be mailed and require small amounts of stool may be an effective solution for collecting fecal samples from children for large scale microbiome studies; however, the quality of sequencing resulting from this is unknown.

Aims: To compare 16s rRNA sequencing results from stool, and also meconium, stored on a FOBC vs. in an eppendorf tube (ET) under different conditions.

Methods: 8 stool samples from children in diapers aged 1 month-2 years and 3 meconium samples were collected and stored as follows: 1) ≤2days at room temperature (RT) in an ET 2) 7 days at -80°C in an ET 3) 3-5 days at RT on a FOBC 4) 7 days at RT on a FOBC and 5) 7 days at -80°C on a FOBC. DNA was extracted and each specimen/condition was sequenced with replicates on the Illumina MiSeq. Overall microbiome structure and taxa distributions were compared between collection method. Alpha diversity (observed, Shannon, Simpson) was compared pairwise between different storage conditions. The Adonis method was used to determine whether the 5 different conditions used for storing the samples were different based on unweighted unifrac distances.

Results: Overall microbiome structure differed between individual stool specimens as expected  (p<10-5), but there was no significant difference between the storage method (p=0.18).  However there was a significant difference between storage methods for meconium (p=0.039). For alpha diversity, when compared to a goal standard of stool in an ET  at RT for <2 days, there was no difference in diversity for FOBCs at 7 days at RT or 7 days at -80°C. Stool stored on FOBCs did tend to have an increase in Firmicutes and a decrease in Proteobacteria compared with ETs.

Conclusion: In stool collected from diapers from young children, there was no significant difference in alpha and beta diversity from stool collected and stored on FOBCs compared with fresh or frozen stool in ETs. There was a significant different in microbiome structure between storage conditions for meconium however. Collection of stool and mailing on fecal occult blood cards may be a low-cost.effective method for large scale population based microbiome studies in children, but not for meconium.



Sara Hajizadeh Barfjani, Jeffrey D. Goldsmith, Alison Cross, Elana Bern, Susanna Y. Huh, Boston Children’s Hospital, Boston, MA, USA  

Background: In infants and young children with feeding difficulties, esophagogastroduodenoscopy (EGD) is increasingly performed to look for reflux or eosinophilic esophagitis, but few studies have examined the prevalence of esophagitis and associated risk factors in this population.  Identification of risk factors associated with esophagitis could help physicians decide which patients should undergo EGD.   

Aim: Among young children with feeding difficulties who underwent EGD, to determine the prevalence of esophagitis and compare the clinical characteristics of children with and without esophagitis.

Methods: We retrospectively reviewed electronic medical records of children who met the following inclusion criteria: initial outpatient gastroenterology visit at Boston Children’s Hospital between January 2014 and December 2015 with an ICD-9 diagnostic billing code for feeding difficulties or dysphagia, and 0 to 36 months old at the time of their first EGD. We excluded subjects with pre-existing enteropathies or congenital GI anomalies. We recorded demographic and clinical data using a data abstraction form. We compared characteristics of children with and without histologic esophagitis using chi-square and t-tests.

Results: Of 171 subjects, 56% were <1 year old. Of 38 (22% of 171) subjects with histologic esophagitis, 37 had eosinophilic and 1 had neutrophilic inflammation. Gross and histologic findings were not well-correlated; 72% of subjects with esophagitis had grossly normal mucosa. Of the 22 subjects (13% of 171) with a history of food allergy, 64% had esophagitis. Food allergy history was present in 8 of 134 (6%) subjects without esophagitis, 10 of 25 (40%) subjects with <20 eosinophils/hpf, and 4 of 12 (33%) subjects with >20 eosinophils/hpf (p<0.0001). Subjects with and without esophagitis did not differ in presenting symptom rates of gastroesophageal reflux or vomiting (63% vs. 65%, p=0.80), food refusal (37% vs. 41%, p=0.68), and choking or gagging with food (37% vs. 45%, p=0.36).  Subjects with esophagitis had lower rates of fussiness as a presenting symptom (5% vs. 20%, p=0.03) but only 2 of 29 children who presented with fussiness had esophagitis.  Subjects with and without esophagitis did not differ in history of past or current proton pump inhibitor (71% vs. 76%, p=0.57) or elemental formula use (32% vs. 19%, p=0.11).

Conclusion: Almost 1 in 4 infants and toddlers under 3 years old who presented with feeding difficulties had esophagitis, usually with increased eosinophils. Among the small number of subjects with food allergy, 64% had esophagitis. These data suggest that clinicians should consider EGD to look for esophagitis in infants and toddlers with feeding difficulties and a history of food allergy.



Tan Teck King, Chang-Ching Wei, An-Chyi Chen, Cheng-Li Lin, Te-Chun Shen, Tsai-Chung Li, Taiwan Society of Pediatric Gastroenterology, Hepatology and Nutrition, Taichung City, Taiwan

Objectives: To systemically investigate the risk of subsequent irritable bowel syndrome (IBS) in children with antecedent allergic diseases in a population-based case-control study in Taiwan.

Methods: We evaluated 11,242 children (age range: 7-18 years) with IBS and 44,968 age and sex-matched control subjects who had been examined between 2000 and 2008. IBS odds ratios (ORs) were calculated for children with antecedent allergic diseases, including allergic conjunctivitis (AC), allergic rhinitis (AR), asthma, atopic dermatitis (AD), urticaria, and food allergy (FA).

Results: Children with antecedent allergic diseases had a greater risk of IBS than did control subjects (p <0.001). Among the 6 evaluated diseases, the highest adjusted OR (aOR) of 1.78 was observed with AR (95% confidence interval [CI], 1.69-1.87), and the lowest aOR of 1.40 was observed with AD (95% CI, 1.2-1.62). With 2 or more allergic diseases, the aORs increased to 2.06 (95% CI, 1.932.19) for all subjects, 2.07 (95% CI, 1.88-2.28) for girls, and 2.18 (95% CI, 2.02-2.35) for children ≥12 years old. The highest aOR of 2.94 (95% CI, 1.35-6.40) was noted when food allergy concurrent with asthma.

Conclusions: Preschoolers with a history of allergic disease had an increased risk of subsequent IBS development upon reaching school age. This risk increased in the presence of concurrent allergic disease and a higher clinical allergy burden.



Vijayalakshmi Kory1, Vaishali Bothra1, Alan Schwartz2, Sue Weides3, James Berman3, Kiranmai Gorla3, Thirumazhisai Gunasekaran3, 1Center for Childrens's Digestive Health, ParkRidge, IL, USA, 2University of Illinois Hospital, Chicago, IL, USA, 3 Advocate Lutheran General Children's Hospital, ParkRidge, IL, USA

EoE in children is associated with many symptoms. Consensus-2011 statement classifies patients into four groups based on predominant symptom; EoE- D, abdominal pain (AP), GERD and failure to thrive. Previously, we showed the features and outcome of EoE-D with and without food impaction (FI) in a small group of 36 patients.1 This study compares a larger group.

Aim: Compare clinical features, endoscopy+ biopsy, and outcomes of 101 EoE – D patients presenting with and without FI.

Methods:  In this retrospective study, patients with EoE-D seen between January2001 and August2015 were stratified into Group I, with FI and Group II, without FI. Physical findings, CBC, esophagram, EGD+ biopsies of the duodenum, antrum, distal and mid=esophagus were captured. Diagnosis of EoE was made as per the consensus guidelines. Treatments included topical or oral steroids, dietary modification, ±PPIs, as per our EoE Clinic protocol. Symptom score for dysphagia, nausea, vomiting, regurgitation, early satiety and heartburn were scored: absent -0, mild -1, and severe-2 except dysphagia had a score up to 3 with FI. Esophageal EGD findings were scored at diagnosis and follow-up.

Results: Patients; Gr I- 30, Gr II- 71, mean age 10.9 and 10.4 yrs. (p=0.27). Clinical features, X-ray, EGD + biopsy findings are given in Tables 1. All patients with FI required endoscopic removal, and strictures, if present, were dilated later. Follow-up: Gr I; mean 1.7 yrs. (range 1/12-8 yrs) and in Gr II 1.1 yrs (1/12–7yrs). Treatments included; fluticasone, diet or combination,±PPI. Symptom improvement; Gr I mean dysphagia score improved from 3 to 1.25 (p<.001) and  Gr II- 1.21 to 0.71 (p<.001) and mean composite score from 3.3 to 1.45 (p<.001) and 2.49 to 1.48 (p<.001). EGD: Gr I, 15/30 (50%) and Gr II 45/71(63.4%) had a follow-up, at 8-12 weeks. Cumulative EGD score improved from 1.8 to 1.7 in Gr I (p=.70) and from 1.6 to 1.5 in Gr II (p=0.08); peak eos. count in Gr. I was 53.2 at diagnosis and 37.5 after treatment (p=.04), and in Gr. II 47.6 and 29.8 (p=.001). Mean eos. count in Gr. I was 42 before treatment and 27.5 after (p=.03), and in Gr. II 39.3 and 22.1 (p<.001). Patients with strictures (8); it stayed open after dilation while the small caliber esophagus was recalcitrant. 1/30 (3.3%) patients in Gr I had recurrence of FI and none in Gr II developed FI. There were no perforations.

Conclusion: There were no significant differences in the clinical features and endoscopic findings and outcome of dysphagia of the two groups.  With treatment there was a significant improvement in the eosinophil count in Group II (without FI) and not in Gr I. Strictures opened up and remained open, while small caliber esophagus was recalcitrant. 1/30 (3.3%) patient in Gr I had recurrence of FI and none in Gr II developed FI. More prospective studies with long-term follow-up are needed to validate this data.

Reference:1. Gunasekaran T. Characterization of Dysphagia Associated EoE in Children with and witho



Presenting Symptoms, Esophagram, EGD and Biopsy findings at Diagnosis:


Gr I Food Impaction (n=30)

Gr II No Food Impaction (n=71)

P value













Mean age






Presenting Symptoms







Associated allergies





























0 .47

GI bleeding












Number of patients who had Esophagram







Gr I Abnormal: 2 stricture, 1 small caliber esophagus,1 spasm.                                                      Gr II Abnormal: 2 stricture,1 Spasm, 3 esophageal reflux, 2 abnormal  swallowing, 3 mucosal irregularity, 1 Esophageal tear

EGD Findings































Biopsy at diagnosis

Peak eosinophil count (mean)

         53       (37.5)

        47.6      (29.8)






Eosinophilic  microabscess






Basal cell hyperplasia




















Thomas Attard1, Mikaela Miller1, Ashwath Kumar2, Chaitanya Pant2, Mike Thomson3, 1Children's Mercy Hospitals & Clinics, Kansas City, MO, USA, 2University of Missouri-Kansas City School of Medicine, Kansas City, MO, USA, 3Sheffield Children's Hospital NHS Foundation Trust, Sheffield, UK        

Background: The risk factors associated with mortality in children with gastrointestinal hemorrhage (GIH) are poorly understood. This is presumably related to the rarity of fatal outcomes limiting the feasibility of prospective studies. GIH in children frequently complicates multisystem chronic illness. The Pediatric Health Information System (PHIS) database collects admission, diagnostic, and treatment data among 44 children's hospitals across the United States (U.S.) and affords an insight on the demographic and clinical characteristics that are associated with mortality in children with GIH.

Methods: The study is a retrospective multi-institutional database analysis. PHIS was interrogated through combined discharge diagnosis (ICD 9, ICD 10), procedure codes (CPT) and pharmacology therapeutic codes (CTC) for children diagnosed with GIH from 2011 to 2015 at the time of admission or complicating their inpatient course.  Demographics, co-morbidities, inpatient course including pharmacotherapy and procedures were analyzed using the generalized linear mixed model framework. The R statistical package was used for analysis.

Results: During the period studied, 61,629 children were diagnosed with GIH through ER (43,254) 24-hour observation unit (3198), ambulatory surgery unit (257) and inpatient service (14,200). Mortality was 0.5% overall (M:F 1.1:1), mean age (SD) was 8.4 (7.4) years for mortality cases, and 6.8 (7.3) years for non-mortality cases. No difference was noted across racial definitions. Median inpatient length of stay in children with GIH who died was 18.5 days (range: 1 to 687 days), compared with 1 day for non-mortalities (range: 1 to 376 days). After adjustment for potential confounding factors, mortality was significantly associated with urban/rural residence (p= 0.007), being higher in children reported to live in rural zip codes (OR 1.65 95% CI, 1.21-2.26 ), and when GIH was not reported at the time of admission compared to complicating inpatient course (OR 1.84 95% CI, 1.47-2.30,P value <0.0001). Mortality was higher in chronic liver disease (CLD) (OR 3.36 95% CI, 2.31-4.88 p < 0.0001) although chronic complex disease was present in nearly all mortality cases (97%). Treatment with proton pump inhibitor therapy (79%), H2 receptor antagonists (43%), erythromycin (9%) and octreotide/vasopressin (18%) was not ubiquitous to all mortalities; octreotide use highly associated with CLD diagnosis 2(1) = 2010.1, p< 0.0001).

Conclusions: This study represents the largest cohort of patients diagnosed with GIH and underscores important differences in children who succumb after the diagnosis of GIH compared to survivors. Early intervention and rapid access to specialized care appears to be important. The occurrence of GIH in children with chronic disease, especially liver disease, appears especially ominous and demands greater vigilance and greater adherence to standard pharmacologic interventions including acid suppression.



Wael N. Sayej, Marina L. Fernandez, Susan Goodine, Connecticut Children's Medical Center, Hartford, CT, USA                   

Background: Celiac disease (CD) is an autoimmune disease that has a prevalence of 1% in the general population in the United States. Esophageal eosinophilia due to gastroesophageal reflux disease (GERD) and Eosinophilic Esophagitis (EoE) have been found to be more prevalent in patients with CD than the general population, but the relationship appears to be coincidental.  In this case series, we sought to characterize our patients who have CD and EoE and evaluate their response to gluten-free diet (GFD).

Methods: We searched our clinical databases for the diagnosis codes for CD and EoE and then extracted patients who had both diagnosis codes diagnosed between 2012 and2015. A total of 13 patients were identified. We then reviewed their laboratory data for positive celiac serologies and histology records to confirm the diagnosis of CD based on published diagnostic criteria (villous blunting and increased intraepithelial lymphocytes). We also reviewed the esophageal histology to confirm the diagnosis of EoE based on the 2011 NASPGHAN consensus guidelines (presence of >15 eos/hpf, basal layer hyperplasia and elongated papilla). We collected patient demographics and clinical data. 

Results: A total of 13 patients were identified to have co-existent CD and EoE (ages 1-15 years, mean 8.5 years, 77% females, and 69% Caucasians, 31% with asthma).  11/13 patients were treated with GFD + PPI and 2/13 were treated with GFD only as their initial therapy. A repeat endoscopy was performed 4-9 months after the initial endoscopy. The esophageal eosinophilia resolved or improved in 9/13 (69%) patients on the GFD ± PPIs and 3/13 (23 %) patients were placed on a dairy-free diet and 1/13 (8%) was placed on SFED, which eventually resolved the eosinophilia. The average peak eosinophil count on initial endoscopy was 43 eos/hpf (range 25-80), post- GFD ± PPIs was 7 eos/hpf (range 0-15) in the responders and 33 eos/hpf (range 25-40) in the non-responders. After adding dairy-free diet (n=3) and SFED (n=1) the peak eosinophil count decreased to an average of 2.5 eos/hpf (range 0-5). 

Conclusion: We present a case series of 13 patients with CD who were found to have esophageal eosinophilia. The majority of patients (69%) responded to GFD ± PPI’s, which may indicate that these patients have GERD, EoE due to wheat, or that the esophageal eosinophilia is directly related to the CD.  The rest of the patients responded to other dietary eliminations (dairy, SFED), which indicates that they likely had EoE.  From this series, we propose a therapeutic algorithm to manage esophageal eosinophilia in patients with CD.



Warapan Nakayuenyongsuk, KT Park, Megan Christofferson, Karl Sylvester, David K. Stevenson, Henry C. Lee, Stanford University School of Medicine, Palo Alto, CA, USA

Introduction: Necrotizing enterocolitis (NEC) in preterm infants of very low birth weight imposes high morbidity and mortality. Fecal calprotectin may serve as a viable non-invasive screening test to detect early signs of NEC, predicting low-grade mucosal inflammation prior to overt signs and symptoms of NEC. However, this likelihood has not been explored, as the traditional send-out ELISA calprotectin assay has limited clinical utility due to the time to result. Unlike the ELISA test, the Quantum Blue assay has a rapid turn-around of 15 minutes with the potential for point-of-care use with results possibly in advance of symptoms. As an exploratory pilot study, we aim to examine normative calprotectin values as measured by the Quantum Blue assay in a high-risk cohort.

Methods:  We are conducting a longitudinal cohort analysis of patients admitted to the Lucile Packard Children’s Hospital Neonatal Intensive Care Unit (NICU). The list of all infants admitted to the NICU is reviewed daily, and patients with a birth weight of <1,500 grams are enrolled. Infants with known bowel pathology are excluded from the study. Stools are collected once daily for 30 days or until postmenstrual age of 32 weeks, whichever is longer. Collected stool samples are tested using Quantum Blue® Calprotectin High Range Rapid Test. Recruitment, testing, and analysis are ongoing.

Results:  Thirty-five patients have been enrolled to date. Two were excluded due to early death, leaving samples from 33 patients for data analysis. Of those 33 patients, the majority are male (n=21, 63.6%) and were delivered via c-section (n=26, 78.8%). Mean birth weight is 999.5±307.5g at mean gestational age 28.4 ±  .7 weeks. Preliminary data show two distinct groups in our cohort: first sample in the first week of life with low calprotectin level (<200 mcg/g) versus high calprotectin level (≥200 mcg/g). Descriptive statistics show that maternal indication for preterm birth (i.e., preeclampsia or eclampsia) is significantly correlated with an elevated calprotectin level in the first week of life. Of note, history of maternal chorioamnionitis is not correlated with higher first calprotectin levels. No patient within this cohort has developed NEC, but further enrollment may prove that this subset of patients with higher initial calprotectin values may be more at risk. Additional analysis is required to determine indicators of increasing calprotectin levels after the first week of life and whether these states may be predictive of NEC. 

Conclusion:  In a cohort of premature, at-risk infants for NEC, there exists a subgroup of infants with high perinatal calprotectin levels in the first week of life correlating with maternal causes for preterm birth. Validation is necessary to determine whether maternal causes for preterm birth as a predictor of NEC is detectable by high calprotectin levels. A rapid calprotectin assay has the potential for point-of-care use in this population.



Yao-Jong Yang1,2 , Bor-Shyang Sheu1,2, Chia-Ling Lu2, 1National Cheng Kung University Hospital, Tianan, Taiwan, 2Institute of Clinical Medicine, College of Medicine, National Cheng Kung University, Tainan, Taiwan

Background: The prevalence and disease severity of H. pylori infection increased from childhood to adulthood. The gastric Lewis antigens served as receptors contribute to the H. pylori colonization. Different inflammatory severity and colonization density were demonstrated on gastric epithelium between children and adults with persistent H. pylori infection. However, acute gastric cytokine responses and Lewis antigens maturations after H. pylori infection are unclear. This study aimed to validate the differences of cytokine responses, Lewis b antigen (Leb) expression, and colonization density between youth and adult primary gastric epithelium cell after H. pylori infection.

Methods: We applied a gHuman stomach fetal epithelium (HSFE) cells h and human gastric epithelial immortalized gGES ]1 gcells to mimic child and adult primary gastric epithelial cells. Each group was challenged with H. pylori at various time periods. The H. pylori colonization density and Lewis antigens expression intensity were measured by flow cytometry. Cytokine expressions, including IL-6 and IL-8, were measured by ELISA.

Results:  After H. pylori challenge, the colonization intensity was significantly higher in GES-1 than in HSFE cells. An earlier achieve full density of colonization in GES-1 but the colonization density slowly increased by a time-dependent manner in HSFE cells. H. pylori infection induced Lewis b antigen (Leb) expression in both GES-1 and HSFE cells. The features of Leb increment were compatible with colonization density in both cells. H. pylori-induced IL-6 and IL-8 expressions were significantly higher in HSFE cells than in GES-1 cells, respectively.

Conclusions:  Leb antigen-mediated gastric H. pylori colonization is an acquisition age-dependent process. The gastric IL-6 and IL-8 responses are different between children and adults after H. pylori infection.



YJ Chang, Chang Gung Memorial Hospital, Kwei-Shan, Taoyuan, Taiwan                        

Background: Diagnosing intestinal strangulation complicating a small bowel obstruction (SBO) remains a considerable challenge in children. Our goal was to evaluate the clinicoradiological parameters to predict the presence of strangulated intestine.

Methods : Medical records were reviewed for 226 pediatric patients operated for acute small bowel obstruction over a 15-year period. The clinical, radiologic findings and operation results were examined. Regression analysis was applied to identify parameters that would predict strangulated SBO.

Results:  Of 226 patients with SBO, 65 patients had intestinal strangulation . In multivariable analysis, four clinical variables corrected with intestinal strangulation and were given one point each towards the clinical score: severe continuous abdominal pain, tachycardia, WBC count >14,500 /mm3, and abdominal distention. The area under the receiver operating characteristic curve was 0.77 (CI, 0.69-0.84),with the optimal cutoff of 2. With score< 2, strangulation rate was 16.5 % (95% CI, 0.11-0.23) vs. 74% (95% CI, 0.59-0.85) with score  ≥ 2, (p<0.001). In patients with clinical score  ≥ 2 combined with the presence of ascites on ultrasound or with the wall thickness and reduced wall contrast enhancement on abdominal computed tomography (CT) had the strong evidence of intestinal strangulation (LR 13.5, 95 % CI, 0.6-0.82, p<0.001 ; LR infinite, 95 % CI, 0.58-0.91, p<0.001) .  

Conclusions:  By combining two more clinical parameter including severe continuous abdominal pain, tachycardia, leukocytosis, and abdominal distention with ascites on US or wall thickness and reduced wall contrast enhancement on CT allowed identification of strangulated SBO.



Young Kim, Yim Hyung Guhn, Kwangju Christian Hospital, Gwangju, Chonnam, South Korea                      

Background: The intussusception recurs in approximately 10 percent of children after successful nonoperative reduction. Multiple recurrences of intussusception may occur in those with idiopathic intussusception and is not necessarily an indication for surgery. 

Study Objectives: The main objective was to determine the clinical and sonographic findings that could be used to predict recurrences of ileocolic intussusceptions in children that had been successfully reduced by enema.

Methods: A retrospective search was performed on 83 children with successful enema reduction of ileocolic intussusception, 3 months to 7 years of age, during a 3.2-year period from January 2013 through April 2016. The clinical, laboratory records and sonographic findings were compared between the recurrence of intussusception group (ROI) and the non-recurrence group (NROI).

Results: Nineteen children (22.9%) of recurrence of intussusception were identified. In 16 cases of 19 ROI, and 26 of 64 NROI, the thickening of terminal ileum closest to ileocolic valve was obtained just after successful enema reduction. Statistical significances were found in terminal ileal wall thickening (median, 9.4 vs. 7.95 mm; p=0.0133), in contrast, not in age (median, 24.5 vs. 19.5 months; p=0.204), gender (male, 68.8% vs. 53.8%, p=0.867), irritability (median, 0.1 vs. 0.7 day; p=0.074), currant jelly stool (median, 0.1 vs. 0.1 day; p=1.0), or C-reactive protein (median, 0.69 vs. 1.83 mg/dL; p=0.908).

Conclusion: Recurrence is associated with the thickening of terminal ileal wall but not clinical or laboratory findings. Given the small numbers of cases, further studies should be considered. We recommend a premeditated measurement of terminal ileal wall after successful reduction of ileocolic intussusception.  





Anthony Anani, Lori Mahajan, Elizabeth Collyer, Katherine Lamparyk, Sarah Worley, Cleveland Clinic, Cleveland, OH, USA        

Background: Anxiety is easily provoked in children undergoing invasive medical procedures including endoscopy. Two prior studies have documented the effectiveness of psychological preparation prior to endoscopy for reduction of anxiety in children. Technology now permits the common place use of hand-held devices (smart phones and tablets) as entertainment by parents and patients during medical appointments. We hypothesized that smart applications (apps) on these devices could be used as possible relaxation and distraction tools to reduce pre-procedural anxiety in children.

Objective(s): a) Compare pre-procedural anxiety in pediatric patients undergoing endoscopy within two groups: Intervention group (play with apps) vs.. control group (no apps). b) Identify areas of greatest anxiety for patients and parents during outpatient endoscopy.

Methods: Prospective randomized control study with patients between the ages of 8-18 yrs undergoing outpatient endoscopy. Patients enrolled on day of procedure and randomized into two groups intervention or control. Patients with prior endoscopy, anxiety disorder, on anxiolytics, history of cardiac/metabolic diseases, neurologically impaired/unable to complete questionnaire or who previously received psychological therapies (CBT) were excluded. Intervention group had access to an electronic tablet preloaded with smart applications- (fruit ninja, Koi pond and balloon animals) and were allowed to play with app for  ≥ 10 minutes prior to procedure. Anxiety measured upon arrival and just prior to procedure;anxiety was measured with vital signs (BP, HR, and RR) and a State Trait Anxiety Inventory for Children (STAI-C). Post endoscopy questionnaire completed by patient and parents. 

Results: 46 patients enrolled, with one dropout. Total of 45 patients: 22 in control group vs. 23 in intervention group. After smart app use for a minimum of  ≥ 10 minutes, those in the intervention group had significantly lower mean post-intervention systolic BP (SBP) than the control group (p 0.017). There was also decrease in other surrogates of anxiety (DBP, HR and STAIC) except for RR, albeit not statistically significant. Majority of the patients and their parents in the intervention group found pre procedural use of the app beneficial (55% and 72% respectively). Post-procedure survey of patients and parents revealed IV placement and waiting for procedure as the two most common causes of pre-procedural anxiety.

Conclusions: Smart applications (apps) significantly reduced pre-procedural anxiety as measured by SBP in children undergoing endoscopy. Other surrogates for anxiety (DBP, HR, and STAI-C) also trended downward with smart app use, albeit not statistically significant. IV placement was the key driver of anxiety for patients. Parents reported waiting while child was in the procedure and observing child undergo anesthesia as major causes of anxiety for them during the procedure.



Anupa Dalal, Anupama Chawla, Denease Francis, Stony Brook University Hospital, Stony Brook, NY, USA

Purpose: Diagnosis of gastroesophageal reflux disease (GERD) in children is challenging, as there is no defined gold standard. GERD is suspected by either the presence of acid reflux documented during esophageal pH monitoring and/or by esophagitis documented by endoscopy. There are two methods of esophageal pH monitoring including 24-hour-pH-multichannel-intraluminal-impedance measurement  (pH-MII) or Bravo pH monitoring. It is unclear how often children with GERD will have both pathological reflux detected esophageal pH monitoring, and histopathological changes of mucosal injury seen on esophageal endoscopy.

Methods: We retrospectively examined clinical characteristics of pediatric patients with suspected GERD evaluated at our institution between January 2009 and July 2014. Subjects on anti-reflux medications or with known esophagitis, secondary to conditions other than GERD, were excluded.

Results: There were a total of 220 patient charts reviewed, of which 134 met inclusion criteria. All of the patients had an endoscopy and a 24-48 hour pH monitoring study. The breakdown of results and clinical features is seen in Table 1. Only  a minority (24/86) of subjects with GERD had both an abnormal endoscopy and pH study, while the majority (41/86) had an abnormal pH study only. Also, we found that 21/86 only had an abnormal endoscopy, however of these subjects, 5/21 were newly diagnosed with eosinophilic esophagitis.

Impression: Children with GERD documented by Bravo/Impedance studies did not have significant changes of esophagitis detected on endoscopy. These two methods of detecting GERD are independent of each other and do not yield consistent results. Further studies must be done to establish a standard of care in diagnosing GERD.



Normal pH Normal EE

Abnormal pH

Abnormal EE

Abnormal pH

Normal EE

Normal pH

Abnormal EE




17.9% (24)

30.6% (41)

15.7% (21)

Age (years)





Gender -Male

35.4% (17/48)

70.8% (17/24)

48.8% (20/41)

50% (10/20)






EE: Endoscopic evaluation

pH: 24-48 hour reflux monitoring



Danielle Usatin, Melissa Fernandes, Isabel E. Allen, Emily R. Perito, James Ostroff, Melvin B. Heyman, University of California, San Francisco, CA, USA                

Objectives:  Endoscopic Retrograde Cholangiopancreatography (ERCP) is increasingly utilized in pediatrics. We hypothesize that ERCP has been applied with acceptable complication rates in children reported in the literature during the past two decades.

Methods: A systematic literature search of MEDLINE, Embase, and Web of Science from January 1995 to January 2016 was conducted for observational studies published in English. Studies reporting ERCP complications in patients <21 years of age without history of liver transplant or cholecystectomy were included. A summary estimate of the proportion of children who experienced any complications following ERCP was derived using a random effects meta-analysis.

Results: Thirty-two studies involving 2612 children and 3566 procedures were included. Subjects’ ages ranged from 3 days to 21 years. Procedures were performed for biliary (54%), pancreatic (38%), and other (8%) non-specific indications; 56% of ERCPs were interventional. Procedural complications included post-ERCP pancreatitis in166 (4.7%), bleeding in 22 (0.6%) and infections in 27 (0.8%). Pooled complication rate was 6% (95% CI, 4%- 8%). Pooled estimate of post-ERCP pancreatitis was 3% (95% CI, 2%-5%), and other complications were 1% (95% CI, 2%-5%). In the subset of articles reporting diagnostic ERCPs performed in neonates with cholestasis the pooled complication rate was 3% (95% CI, 1%-7%). Available data limited the ability to report on differences between pediatric-trained and other endoscopists.

Conclusions: Complications associated with pediatric including neonatal ERCP range widely in severity and are reported inconsistently. Our review suggests 6% of children undergoing ERCP have complications, comparable with rates reported in adults undergoing ERCP. Societal guidelines should be established for defining and reporting timing, and severity of adverse events. Further studies using systematic and standardized methodologies are needed to determine the frequency and risk factors for ERCP related complications.



Joshua Carroll, David M. Troendle, Nandini Channabasappa, Bradley A. Barth, University of Texas - Southwestern Medical Center, Children’s Health – Children’s Medical Center Dallas, Dallas, TX, USA  

Introduction: Through the scope (TTS) clips have been used successfully in both adult and pediatric populations to provide mechanical tamponade in cases of gastrointestinal bleeding. Positioning and placement of the TTS clips can be challenging in certain parts of the GI tract. The safety and effectiveness of Over The Scope Clip (OTSC) in gastrointestinal bleeding has been demonstrated in adult studies; however, there are no pediatric reports demonstrating their use or effectiveness. This report is the first to detail the clinical safety and effectiveness of OTSC in the pediatric population

Methods: A bleeding registry was queried for patients where OTSC were utilized in an attempt to achieve hemostasis. Charts for these cases were then reviewed retrospectively. Data collected included age, weight, indications, rates of technical success, need for re-intervention, and complications from the procedure. All lesions were classified utilizing the Forrest classification. Technical success was defined as successful clip positioning and deployment on first attempt. Immediate hemostasis was defined as achieving hemostasis without need for utilization of other modalities during the same endoscopic session. Need for re-intervention was defined as any need for re-intervention during the follow-up period.

Results: 11 cases of OTSC utilization for hemostasis were identified in 10 unique patients and were included for analyses. Procedures were performed between 11/2014 and 5/2016.  The median age at intervention was 14.7 years (range 3.9– 6.8 years). The median weight was 39 kg (range 17.4–85.8 kg). Table 1 summarizes patient and procedural characteristics. Technical success and immediate hemostasis was achieved in all cases and there were no complications. Both patients with anastomotic ulcerations required further interventions. The first (Patient 9) received additional iron infusions which were slowly spaced out following the procedure and his Hgb has remained normal with no further interventions for 163 days. The second (Patient 10) required a second endoscopy to treat additional anastomotic ulcer sites. At repeat endoscopy, the initially placed OTSC remained in place and there was no evidence of ongoing bleeding from this site. A separate site was treated during that session. This patient continues to require transfusions after her second procedure.

Conclusions: We report the first series demonstrating the safety and short-term effectiveness of the OTSC in the pediatric population for acute gastrointestinal bleeding throughout the GI tract. In our experience, it is extremely effective for non-anatomic ulcers and polypectomy bleeds even when other hemostatic techniques have failed. OTSC may be less effective in the setting of anastomotic ulcerations, reaffirming the refractory nature of this type of GI bleed although further work is needed to clarify long-term benefits.



Josh Cousin1, Douglas S. Fishman1, Jose R. Rodriguez1, Surya P. Rednam1, Seth S. Septer2, Deborah Schady1, 1Texas Childrens Hospital-Baylor College of Medicine, Houston, TX, USA, 2Children's Mercy Hospital, Kansas City, MO, USA           

Introduction:  Limited pediatric data is available regarding the removal of large intestinal polyps in children. We evaluated the techniques used to remove polyps ≥15 mm in consecutive pediatric patients from 2006-2015.

Methods: We performed a retrospective single center study using a pathology database paired with endoscopic documentation software from 2001 to2015.

Polyp removal technique was divided into surgical sub-divisions (rigid anoscopy, laparoscopic,  open and other) or endoscopic (snare, snare with pre-clip, snare with post-clip, snare with post-cautery, and other). Polyp type, size (with and without stalk), location and adverse events were recorded. SPSS 23 was used to calculate means, interquartile ratios and differences between groups. Institutional Review Board approval was obtained for this analysis.

Results: Of 648 polyp related procedures (585 unique patients) identified, 113 procedures yielded at least one polyp ≥15 mm from 105 unique patients. Polyp size in this group was 2.18 cm and IQR (1.68-2.5) with a maximum of 5.5 cm and 2.35 and IQR (1.7-2.5) when stalk size was included.  There were 11 small intestinal polyps, 83 left-sided colon polyps, 18 right-sided polyps and one colon NOS. 20 patients had more than one large polyp removed at the same session. 86% (88 of 102) were juvenile polyps while polyposis syndromes (1 FAP, 15  JPC, 14 Peutz-Jeghers) accounted for only 26.5% of the patients. Additionally, there was one tubulovillous adenoma, but no advanced malignancies. PJS polyps were larger than JPC (p=.036) polyps.

The mean age at time of removal was 6.5 years (range 6 months to 17 years). There was no difference between patient age and size of polyp=   (p .373). A total of 93 patients had some form of endoscopic snare cautery, standard snare (n=72)  compared to snare with other techniques     (n=21) including hemostatic clip placement, “piecemeal” or bipolar cautery. 19 patients had surgical therapy, including 10 bowel resections and 6 ligations of rectal polyps. There was a difference between size of polyps removed endoscopically vs. surgery (p=.028). Significant adverse events (n=4) included one equipment failure that led to repeat procedure, 2 patients had repeat endoscopies due to size or atypical location, and one patient had bradycardia that resolved with medication.

Conclusion: Large polyp removal in pediatric patients is safe and the majority are removed endoscopically with snare cautery. Surgical intervention was typically for intussusception or obstruction requiring bowel resection with polyp removal.



Harveen K. Singh, Geoffrey D. Withers, L.C.Ee, Lady Cilento Children's Hospital, Brisbane, Queensland, Australia                   

Introduction:  Quality indicators for colonoscopy in adults are mainly driven by colorectal cancer screening, and include cecal intubation and adenoma detection rates. Cecal intubation rates of >90% is recommended in adults by most societies and colleges internationally. In contrast, colorectal cancer is rare in children so colonoscopy is predominantly diagnostic. Common indicators for pediatric colonoscopy include investigating for inflammatory bowel disease (IBD), diarrhoea or abdominal pain. In these conditions, ileal intubation is strongly recommended as it optimizes diagnostic yield. There is, however, a paucity of data on quality indicators for pediatric colonoscopy, and it remains unclear whether high rates of cecal and ileal intubation is achievable in pediatrics.

Aims:  This study was undertaken to audit all colonoscopies performed in a tertiary pediatric center to examine for the clinical indications for procedure, completion rates to cecum and ileum, and rate of significant findings.

Methods: Retrospective review of all colonoscopies from November 2011 to the end of October 2015 was performed. The ORMIS theatre management database was used to identify patients having colonoscopy using ICD-10 codes 32090 (colonoscopy) and 32087 (colonoscopy ± polypectomy). Patients having intentional flexible sigmoidoscopy were excluded from further analysis although incorrectly coded patients who proceeded to total colonoscopy were included. Patient demographics, indication for procedure, presence of trainee, quality of bowel preparation, extent of colonoscopy and confirmation of location, reasons for incomplete procedure, diagnostic findings, and complications were noted.

Results:  652 patients were identified as having had or intended to have total colonoscopy after exclusion of incorrectly coded patients. Median age of patients was 13.0 (range 0.4-18.2) years, with 53% male. The most common indications for colonoscopy were IBD review (57.9%, 378/652), rectal bleeding (10%, 68/652), abdominal pain (10%, 68/652), and diarrhea (8.6%, 56/652). All patients had procedures under general anesthesia. Trainees performed 69.8% (452/652) of procedures. Quality of bowel preparation was mentioned in 62.9% (410/652), of which 21.9% (90/410) were considered inadequate. Cecal intubation rate was 96.3% (628/652) and ileal intubation 92.4% (603/652). Photographs and/or biopsies were used to confirm extent of procedure in 99.2% of patients. Factors predicting success of ileal intubation include quality of bowel preparation and patient age. Normal histology was noted in 61.8% (403/652) of colonoscopies. 37 (5.6%) patients had polypectomy; most were juvenile polyps (54%, 20/37). No perforations occurred but three patients had hematoma, which were managed expectantly.

Conclusion:  High rates (≥90%)of cecal and ileal intubation are achievable in pediatric colonoscopy. Ileal intubation should be considered a quality indicator in diagnostic colonoscopy in pediatrics.



Ilse Julia Broekaert1, Joerg Jahnel2, Marta Tavares3, Nicolette Moes4, Hubert van der Doef5, Christos Tzivinikos6 1Cologne University Children's Hospital, Faculty of Medicine, Cologne, NRW, 2Medical University of Graz, Graz, Styria, Austria, 3University Hospital Porto, Porto, Portugal, 4University Medical Center, Antwerp, Belgium, 5University Medical Center Groningen, Groningen, Netherlands, 6Alder Hey Children's Hospital, Liverpool, UK   

Objectives: Endoscopy training is an essential part of pediatric gastroenterology, hepatology and nutrition (PGHN) fellowship as specified in the ESPGHAN training syllabus. The aim of this study was to evaluate the endoscopy training among fellows and young professionals in PGHN. The recently published ESPGHAN syllabus suggests a minimum of 100 esophagogastroduodenoscopies (EGDs) and 50 colonoscopies for certification (D’Antiga et al., 2014).

Methods: 84 PGHN fellows participated in an electronic survey called by ESPGHAN between February 2014 and September2015. The survey comprised 32 questions on general information, number of endoscopies performed, specific endoscopic procedures, supervision and certification, and endoscopy training.

Results: Among 84 participants 28 (33%) have already finished their training and 42 (50%) are still in training. 53 fellows (63%) reported to be enrolled in an official PGHN fellowship program leading to a subspecialty certification. 32 (38%) devote their entire time to PGHN training and 34 (40%) between 50 and 99% of their time. 66 PGHN fellows (79%) are trained in endoscopy during their fellowship. Of all fellows, 29 (35%) are trained by an adult gastroenterologist and 6 (7%) by surgeons. 30 (36%) follow the ESPGHAN syllabus. Concerning the numbers of endoscopic procedures, PGHN fellows have completed 207 EGDs, 67 colonoscopies, 11 polypectomies, 10 variceal bandings and 20 PEG changes/ insertions on average. The terminal ileum is intubated in 29% most of the time (>90%). 63 fellows (75%) enjoy continuous supervision, 65 fellows (77%) keep an endoscopy logbook, and 28 (33%) have formal assessments (paper or online) during and 47 (56%) at the end of their training.

During their training 54 fellows (64%) have attended basic skills endoscopy courses and 43 fellows (51%) have completed endoscopy simulator trainings. 79 fellows (94%) wish participation in future ESPGHAN endoscopy summer schools  and 75 fellows (89%) would like to attend basic endoscopy skills courses. Fellows feel that their upper GI endoscopy training will allow practicing as consultant in 86% and their colonoscopy training in 67%. 59 fellows (70%) would like ESPGHAN to be responsible for the accreditation of endoscopy centers.

Conclusions:  This survey shows that endoscopy training differs among fellows in Europe regarding accomplished procedures, the training program including supervision and certification, and specific endoscopy courses. Only 36% have followed the ESPGHAN training syllabus and only 86%, respectively 67%, feel skilled enough to perform EGDs and colonoscopies when practicing as a consultant. We encourage all European GI centers to follow the ESPGHAN training syllabus to harmonize endoscopy training during PGHN fellowship throughout Europe, eventually leading to better endoscopy skills of young consultants. 



I. Irastorza1, R. Barrena2, M. Legarda3, C. Tutau3, 1Hospital Universitario Cruces. UPV-EHU, Barakaldo, Basque Country, Spain, 2Clínica Zorrotzaurre, Bilbao, Basque Country, Spain, 3Hospital Universitario Cruces. UPV-EHU, Barakaldo, Basque Country, Spain

Introduction: Inadequate colonic cleansing is frequent in pediatric patients undergoing colonoscopy and is the main reason to admit the patient to hospital for bowel preparation. Hospital admission significantly increases the cost of the procedure and causes family discomfort. The aim of the study was to assess the effectiveness of a regime of outpatient bowel preparation with macrogol 3350 in patients older than 2 years.

Material and methods: 390 children, 2 to 14 years old (mean age 8.6 years) were given macrogol 3350 with electrolytes and ascorbate for bowel preparation the day before colonoscopy with a weight-adjusted dose of 2 g per kg. If the child failed to take the medication at home, hospital admission was arranged in order to administrate the bowel preparation by a nasogastric tube (NGT). Efficacy in colon preparation, side effects and the need for admission to complete the bowel preparation were assessed.

Results: 345 children (88%) completed the bowel preparation at home. In 295 (76%) colonic cleansing was adequate, in 44 (11%) it was incomplete but allowed us to explore the entire colon and in 5 (1%) colonic preparation was inadequate and colonoscopy was cancelled. 47 (12%) children failed to complete bowel preparation at home and were admitted to hospital (16 for vomiting and 31 for inability to drink the medication): in 26 (7%) children bowel preparation through NGT was successful, 15 (4%) children vomited the medication given by NGT and in 5 (1%) patients, NGT administration of the bowel preparation had to be stopped due to intense abdominal cramping. In total, 39 (10%) children reported cramping abdominal pain and 21 (5%) children reported vomiting. Overall, colonic cleansing at home was successful in 87% of children.

Conclusions: Outpatient bowel preparation with macrogol 3350 (2 g per kg) is successful in most cases and ensures adequate colonic cleansing resulting in a significant reduction of costs and family discomfort.



Jacob Mark, Robert Kramer, Cara Mack, Children’s Hospital Colorado, Aurora, CO, USA

Background: Endoscopic retrograde cholangiopancreatography (ERCP) is used for a variety of biliary and pancreatic indications in pediatric patients including biliary strictures, common bile duct stones, biliary leaks, pancreatic duct strictures and others.  Biliary stent placement is often used for biliary strictures and other obstructive lesions to maintain bile flow. Multiple plastic stents are used commonly in pediatric patients but may not be effective in some cases due to migration or obstruction of the stents. Self-expanding metal stents (SEMS) have been used more recently by adult gastroenterologists, for both malignant and benign lesions but their use and placement in adolescents and children has not been well described by pediatric endoscopists.  Case Presentation: We present a series of two adolescents and one young adult with complex medical needs who had biliary SEMS placed for different indications. Patient 1 is a15-year-old, 53-kg adolescent who had liver transplant for CF related liver disease and refractory post transplant anastomotic stricture that did not resolve with two rounds of multiple plastic stenting. He had improvement in bile duct diameter and GGT after 10 mm x 8 cm SEM placement (Wallflex, covered metal biliary stent, Boston Scientific, Boston, MA) for 7 weeks which was sustained at 3.5 months after placement. Patient 2 is 18-year-old, 7- kg adolescent with autoimmune hepatitis who had liver biopsy complicated by gallbladder perforation bile peritonitis. She had persistent bile leak after partial cholecystectomy, which was technically difficult because of surrounding inflammation, and did not resolve with sphincterotomy and multiple plastic stents placed across the cystic duct. Bile leak resolved with covered SEM placement 8 mm x 8 cm across the cystic duct for 6 weeks and she has had no symptoms of bile leak recurrence 4 months after last stent removed. Patient 3 is a 24-year-old, 36-kg young woman with Rett syndrome and multiple pulmonary and neurologic comorbidities who had obstructive choledocholithiasis 2 years after cholecystectomy. She continued to have multiple biliary stones and signs of obstructive choledocholithiasis after multiple ERCPs with balloon sweeps, sphincterotomy and multiple plastic stenting so 10 mm 6.0 cm SEM was placed with resolution of biliary obstruction after 8 weeks. In all cases, the stents were well tolerated and no adverse events were encountered. Removal was performed endoscopically using a raptor forcep (US Endoscopy, Mentor, OH) without difficulty.

Conclusion: SEMs may be considered by pediatric endoscopists for patients with benign obstructive lesions or perforations of the common bile duct who are not responding to plastic stents. Additional research is needed, however, for more comprehensive evaluation of the efficacy and potential adverse events in pediatric patients undergoing SEM therapy. 



Jamal Kriem, Riad Rahhal, University of Iowa, Iowa City, IO, USA                     

Background: Esophageal food impaction is one of the conditions that often requires immediate attention including urgent endoscopy. Adult- based guidelines support the use of the extraction (pull) technique but allow the consideration of the advancement (push) technique with caution due to the high probability of esophageal pathology and risk of perforation. The NASPGHAN guidelines mention the use of gentle advancement for disimpaction but elaborate that the use of this technique in children has not been studied.

Hypothesis: The push technique is safe and effective in the treatment of pediatric esophageal food impactions.

Methods: This was a retrospective cohort study of all pediatric patients presenting with esophageal food impactions to a pediatric tertiary care center from 2003 to 2016. Procedures were identified using billing records.

Results: Two hundred and forty-two procedures were identified based billing codes for esophageal foreign body removal. Forty procedures were for treatment of esophageal food impaction in a total of 24 patients (range: 1-4 procedures per patient). The most common underlying diagnoses were eosinophilic esophagitis (42%) and history of tracheoesophageal fistula (38%). The cohort had a median age of 8.5 years and median weight of 35.2 kg. Initial endoscopic disimpaction methods include 21 push technique and 19 pull technique attempts with success rates of 62% and 68% respectively (p=0.67). Unsuccessful attempts using one technique were successfully accomplished using the other technique. All patients were discharged within 24 hours of the procedure, except for one patient who was transferred from another hospital with an esophageal perforation secondary to a failed disimpaction. The perforation was managed conservatively. No procedure-related complications were reported at our center. The two groups of patients (managed by the two disimpaction techniques) did not differ in age, weight, gender, presenting symptoms, type of anesthesia used or underlying diagnoses except that patients with known fixed esophageal strictures (n=4) were managed using the pull technique only.

Conclusion: This study shows that the push technique is as safe and effective as the pull technique in managing esophageal food impactions in pediatric patients in the absence of known fixed esophageal strictures.



James Brief, Anupama Chawla, Jeffrey Morganstern, Stony Brook Children's Hospital, Stony Brook, NY, USA   

Background: Colonoscopies can be a source of anxiety for patients and their families. From a grant provided by NASPGHAN, we developed software (an “app”) that familiarizes patients with the procedure, answers commonly-asked questions, informs patients of when and where to report for their colonoscopy and guides patients through the colonoscopy prep process, providing real-time instructions about which medications to take, when to take them and how they should be prepared.

Methods: 46 patients aged 5-18 were randomized to receive either written or software prep instructions. Prep quality was measured with the Boston Scoring Scale. The number of calls to the gastroenterology service were recorded. Patient arrival time was recorded on the day of their procedure. A questionnaire was given to patients on the day of the colonoscopy.

Results: App users had superior mean Boston scores of 9.80 versus controls’ 7.96 (p=  0.014). Although not statistically significant, 10/20 (50%) app users had improved knowledge of the colonoscopy versus 8/22 (36.4%) controls (p=0.37). App users made fewer phone calls to the GI service than controls (6 vs. 2), although this difference also did not reach statistical significance (p=0.27). There was no difference between arrival times at the endoscopy suite between app users and controls (p= 0.56). 

Conclusion: App users had significantly better quality preps than control subjects. While results showed a trend towards app patients feeling better informed and knowledgeable about the colonoscopy prep, and requiring less physician guidance, these results were not statistically significant. App and control subjects arrived at the endoscopy suite at nearly the same time. We anticipate that future studies with greater numbers of subjects will reach statistical significance for these measures.



Javier Monagas, Lauren Del Bosque, James Noel, R. Adam Noel, Baylor College of Medicine, San Antonio, Texas, USA             

Background: Most pediatric gastroenterologists use the pull technique for percutaneous endoscopic gastrostomy tube placement (PEG). Gastropexy is a new technique for endoscopic gastrostomy tube placement. T-fasteners are placed through the skin into the stomach, attaching the stomach to the abdominal wall, visualized directly by endoscopy. Serial dilators are used to create a gastrostomy tract, allowing a low profile gastrostomy (GT) or gastrojejunostomy tube (GJT) to be placed. There are no published studies that compare the gastropexy procedure outcomes to the standard PEG procedure in pediatrics.

Objective: To study the safety and outcomes following gastropexy vs. PEG.

Methods: Gastropexy and PEG patient charts were compared for immediate complications(perforation, bleeding, and infection), pain and long term complications (feeding problems, pain, tube dislodgement, infection, bleeding, granulation, readmission, and death) in the three months following procedure.

Inclusion criteria: Pediatric patients requiring gastrostomy placement for enteral nutrition. Subjects were from 0-17 years of age that underwent a procedure in 2014 to 2015.

Results: 13 subjects who underwent gastropexy were between the ages of 1-16 years (avg. 6.6 years), 6 males, with average weight of 23 kg. Fourteen subjects who underwent PEG placement ranged from 0.16–9.5 years of age (avg. 1.5 years), 7 males, with average weight 7.2 kg. There were no immediate short-term complications in either group. Long-term complications were assigned a score according to the severity values, ranging from 0 to 7. The average complication score for both groups was 3, with a median and mode of 2. The procedure length averaged 26 minutes (min) in the gastropexy group compared to 10 min in the PEG group. Average anesthesia time was similar- 53 min in gastropexy group versus 47 min in PEG. Pain treatment for mild, moderate and severe pain was similar in both groups except Toradol that was only used by the gastropexy group. Pain treatment was shorter in the gastropexy group (avg. 1.7 days) vs. PEG group (avg. 3.5 days). Time to start clears was <24 hours in 11/12 subjects for the gastropexy group and in 10/11 subjects for the PEG group. Time to start feeds was <36 hours in 8/11 subjects for the gastropexy group and 12/14 subjects for the PEG group.

Discussion: Our results show that the gastropexy technique is a safe and effective alternative for placement of GT and GJT. The major advantage is that patients who underwent gastropexy did not require a second procedure compared to PEG patients for low profile tube conversion or subsequent placement of a GJT. Disadvantages include longer procedure time for the gastropexy group likely due to the learning curve and differences between operators, but the last 5 procedures averaged 17 minutes.



Siala Nadia, Jeridi Yasmine, Mohamed Lamouchi, Ouerda Hayfa, Azzabi Ons, Maherzi Ahmed, Mongi Slim Hospital, Sidi Daoud, La Marsa, Tunisia

Background: The endoscopic treatment of esophageal stenosis seems to be the most frequently used strategy in children. Improvement in endoscopes and techniques have led to the increase in the number of patients who are conservatively treated with endoscopic dilation rather than surgical treatment. This is a report of our experience with esophageal dilation, its indications, methods and results in children.

Methods: Retrospective study of children  admitted in our pediatric endoscopy unit for esophageal dilation between November 2007 and March 2016.

Resuts: 39 admissions were registered. There were 24 males and 15 females with a mean age of 4 years. Esophageal atresia anastomotic stenosis (n=2, 25.6%) and post-corrosive esophagitis(n=0, 25.6%) are the most frequent types of cicatricial esophageal stenosis. The other indications were peptic stenosis (n=, 15.3%), achalasia (n=,12,82%), congenital stenosis(n=, 7.6%), herpetic esophagitis (n=, 2.5%) and unidentified causes (n=2, 5.1%). The average number of sessions to achieve adequate dilation was 4 per patient with extremes from 1 to 13. Different dilators were used: in 86 cases (54%) we used the balloon dilators, whereas in 37 (23.2%), the Savary-Giliard bougie. Balloon followed by Savary-Giliard bougie dilation had been used in 25 cases (15.7%). In 11 sessions, the dilation method had not been mentioned. Dilation was successful in all cases except one.

Conclusion: The conservative treatment of esophageal stenosis rather than surgery is a well-known strategy for children. Young patients can be treated effectively and safely only by endoscopic dilation.



Nicholas Carman<