Etiology of Esophageal Food Impactions in Children

Christine Waasdorp Hurtado, Glenn T. Furuta, and yRobert E. Kramer

Objective: The aim of the study was to measure clinicopathological features
of children presenting to a tertiary care emergency department with esophageal food impaction.
Materials and Methods: A retrospective chart review of children with
esophageal food impaction seen between January 1, 2005 and June 30, 2009,
including all patients from age 1 month to 18 years with esophageal food
impaction at a pediatric emergency department, was performed.
Results: Initial screening of International Classification of Disease, 9th
Revision, discharge diagnosis identified 698 children with an esophageal
foreign body. Of this group, 72 esophageal food impaction events were
identified in 65 children (69% boys), 49 of whom required endoscopic
intervention. Endoscopic appearances of the esophageal mucosa were
abnormal in 40 (82%), revealing evidence of esophagitis (55%) or
stricture (27%). Twenty-four of the subjects had biopsies taken at the
time of endoscopy. Inflammation, described as increased eosinophils,
basilar hyperplasia, rete peg elongation, and/or microabscess, was present
in 76% of mucosal samples. Follow-up endoscopy in 12 children identified
an etiology in 9, five of whom were found to have eosinophilic esophagitis.
Conclusions: The majority of children with esophageal food impaction
who underwent endoscopic evaluation and biopsy have an underlying
potentially treatable cause. We therefore recommend that all of the
children with esophageal food impaction have mucosal biopsies at the
time of endoscopic disimpaction with appropriate follow-up to allow for
diagnosis and management of the underlying etiology.
Key Words: endoscopy, eosinophilic esophagitis, esophageal food

(JPGN 2011;52: 43–46)


lthough a number of studies have documented the epidemiology and potential etiologies of esophageal food impaction
(EFI) in adults (1–4), few studies have examined these issues in
children (1,3,5–8). Peptic and allergic inflammation, fixed narrowing from congenital rings or Schatzki rings, and dysmotility have
been identified in adults with EFI and suggested in children. Cheung
et al (5) studied 12 children with EFI seen during the course of a
decade, providing the largest study to date. Although no radio-

Received March 22, 2010; accepted April 30, 2010.
From the Gastrointestinal Eosinophilic Diseases Program, and the ySection
Pediatric Gastroenterology, Hepatology, and Nutrition, Department of
Pediatrics, Digestive Health Institute, University of Colorado School of
Medicine, The Children’s Hospital, Aurora, CO.
Address correspondence and reprint requests to Christine Waasdorp
Hurtado, MD, University of Colorado Health Science Center, 13123
East 16th Ave, B290, Aurora, CO 80045 (e-mail: waasdorp.christine@
Dr Furuta contributed equally as senior author.
Glenn Furuta is a consultant to Meritage Pharma. The other authors report no
conflicts of interest.
Copyright # 2010 by European Society for Pediatric Gastroenterology,
Hepatology, and Nutrition and North American Society for Pediatric
Gastroenterology, Hepatology, and Nutrition
DOI: 10.1097/MPG.0b013e3181e67072


Volume 52, Number 1, January 2011

graphic or endoscopic abnormalities were observed, 5 children
demonstrated esophageal inflammation (5). More recently, Vicente
et al (8) reported that 8 children with EFI had histological evidence
of esophageal inflammation.
An increasing body of literature has documented the association of EFI with eosinophilic esophagitis (EoE) (2,3). Our recent
study documented that 56% of adults with food impaction demonstrated clinicopathological features of EoE (2). Given these data, we
hypothesized that EoE was a common underlying cause of EFI in
children. The aim of the present study was to identify clinicopathologic features of children with EFI presenting to an emergency
department at a tertiary care referral center. The secondary aim was
to identify children with clinicopathological features of EoE within
the population of children with EFI.

Billing records of children ages 1 month to 18 years who
were cared for at The Children’s Hospital in Denver between
January 1, 2005 and June 30, 2009 were screened using the
International Classification of Disease, 9th Revision, Clinical
Modification discharge code 935.1 (esophageal foreign body). A
data collection form was used to record data by a single investigator
(C.W.H.) focusing on demographic, radiologic, endoscopic, and
histological features of children with EFI. Included in the data
collection were the date of food impaction, history of allergic
disease, medications at the time of the evaluation, and family
history of esophageal disease. Random review to assess for variability was completed by a second investigator (R.K.). Records that
identified a nonfood esophageal foreign body were excluded from
this study. Only the first admission for EFI was included in the
analysis for children noted to have >1. Recurrent EFI were not
included to avoid skew by 2 patients with multiple events. The
present study was conducted with the approval of the local institutional review board.
Statistical methods include descriptive statistics including
derivation of means, medians with interquartile ranges, and ranges
for continuous data. The x2 goodness of fit was used to evaluate
differences in seasonal variation. Statistical significance was
assumed when the P value was <0.05. Missing data were noted
with no interpolation. JavaStat and GraphPad were used for completion of the statistical review (9,10).

Initial review of the billing records identified 698 patients
with an esophageal foreign body (Fig. 1). Of these, 65 children were
identified with 72 EFI events, with several children with recurrent
EFI and 2 children having 3 impactions (Fig. 1). Sixteen did not
require endoscopic disimpaction but were similar to the 49 who did,
with a mean age of 9.9 years, 62% boys, with similar allergies, but a
slightly different racial composition (Fig. 1). Demographic analysis
of the 49 requiring endoscopic disimpaction demonstrated that boys
more frequently experienced EFI, and that 20% of subjects had
family members with a history of atopy and 6% with a history of
esophageal narrowing or impaction. Twenty-two children (44%)


Copyright 2010 by ESPGHAN and NASPGHAN. Unauthorized reproduction of this article is prohibited.

although not significantly more. 8 had evidence of significant eosinophilic inflammation by histology. Number 1.2 distal biopsies. with 6 demonstrating strictures. and/or microabscess. At initial endoscopy. rete peg elongation. Two of the 3 demonstrated eosinophilic 44 Volume 52. 2). however. Forty-nine of the children with EFI required endoscopic intervention due to lack of spontaneous resolution of EFI and/or intolerance of secretions. 6 of the 12. The esophageal mucosa appeared abnormal in 40 children (82%) with a wide range of different endoscopic features including evidence of inflammation and strictures (Table 3). including 2 of the subjects with stricture. . One patient had fibrosis with eosinophils. Degranulation was noted on the pathology report in 11 of the 24 children with biopsies. Thirteen children were identified with strictures at the time of disimpaction. Several other children were requested to follow up. The food bolus was removed with a combination of endoscopic Roth net and rat tooth forceps in the majority of subjects. documentation of follow-up was found only for the 12 stated above. Summary of demographic information from children with EFI Esophageal food impaction events 72 5 subjects with multiple EFI First esophageal food impactions 65 16 spontaneous resolution Endoscopic disimpaction 49 25 without biopsy Mucosal biopsy obtained 24 3 Normal  Mean age (range) Race (%) White/non-Hispanic White/Hispanic African American Asian Not reported Sex (%) Male Female History of atopy (%) Allergies Environment Food Medications Animals Asthma Eczema Family history (%) Atopy Strictures Food impaction GERD Medication (%) None PPI or H2 receptor antagonist Asthma Inhaled steroid Allergy Topical steroid Montelukast inhibitor Other Subjects requiring endoscopy n ¼ 49 Spontaneous resolution n ¼ 16 10.Hurtado et al JPGN Esophageal foreign body 698 21 Abnormal histology 12 with follow-up endoscopy 4 with history of TEF 3 Normal appearance 9 Abnormal appearance 5 Eosinophils > 20 per hpf 4 Normal biopsy 3 No biopsy FIGURE 1. PPI ¼ proton pump inhibitor. 5 of whom did not have biopsies at the time of the first endoscopy. 9 were visually abnormal during endoscopy. EFI occurred more commonly in the summer and winter. Four of the 6 with strictures on initial endoscopy were children with a history of TEF/EA repair. January 2011 TABLE 1.6 proximal biopsies and 2. with the remaining 8 having no history of esophageal surgery or previous stricture. Meat was the most common food impacted (Table 2). were taking no medications at the time of their disimpaction (Table 1). infiltrate of >20 eos/hpf with degranulation and the third was noted to have significant eosinophils. The 15 children with >20 eos/hpf included 9 with elevated eosinophil counts seen in both proximal and distal esophagus and 6 with either proximal or distal elevations (Table 4). basilar hyperplasia. Of the 10 children taking inhaled or nasal topical steroid. had evidence of inflammation by histology. Eleven of the 12 had visually abnormal mucosa at the time of the initial endoscopy. GERD ¼ gastroesophageal reflux disease. neutrophilic infiltrate.4 y (7 mo–17 y) 42 (86) 2 (4) 1 (2) 0 4 (8) 9. Histology evaluation completed by standard protocol in the same laboratory using hematoxylin and eosin stain identified inflammation in 20 subjects described as eosinophilic infiltrate. and prevalence increased during the course of the study (Fig. On follow-up endoscopy 5 continued to show www. 15 demonstrated >20 eosinophils per high-power field (eos/hpf) and 3 children had <20 eos/hpf (Table 4). Degranulation was characterized as the presence of extracellular evidence of pink granules in the squamous epithelia. in 24 (49%) patients and were abnormal in 21 (88%) (Table 4). with only 4 subjects having biopsies taken from only 1 location (Table 4). Three children had biopsy results only identified as esophagus. Unauthorized reproduction of this article is prohibited. Subject identification and selection with identification of key study outcomes. Twelve children had follow-up evaluation at the request of the attending physician. none of the patients who had EFI without biopsy went on to have subsequent EFI with an endoscopy and associated biopsy at our institution.jpgn. Mucosal biopsies were obtained. Of the children with eosinophilic infiltrate. at the discretion of the attending physician.org Copyright 2010 by ESPGHAN and NASPGHAN. with none receiving simultaneous disimpaction and dilation. Twenty of the 24 subjects had both proximal and distal esophageal biopsies taken with an average of 2. and 2 did not have biopsies taken.9 y (23 mo–18 y) 11 (69) 2 (13) 1 (6) 1 (6) 2 (13) 34 15 24 21 14 11 5 1 11 1 (69) (31) (49) (43) (29) (22) (8) (2) (22) (2) 10 3 3 2 (20) (6) (6) (4) 22 (45) 4 (8) 9 8 7 2 2 6 (18) (16) (14) (4) (4) (12) 10 6 6 5 (62) (38) (38) (31) 0 3 (19) 2 (13) 0 3 (19) 0 NA 10 (63) 1 (6) 3 (19) 1 (6) 2 (13) 0 1 (6) 1 (6) EFI ¼ esophageal food impaction. with 28 of the EFI removed by a pediatric gastroenterologist and 21 by a pediatric surgeon. Five of the 13 identified strictures were related to repair of tracheoesophageal fistula/esophageal atresia (TEF/EA). NA ¼ not available. The removal of foreign bodies alternated days between gastroenterology and surgery services. To our knowledge.

making the determination of disease at the time of 45 Copyright 2010 by ESPGHAN and NASPGHAN. however. others had clinicopathological features suggestive of EoE such as longitudinal furrows. Previous studies examining clinical features of EFI have shown a male predominance similar to that seen in our population (7. Finally. The present study also identified an underlying etiology in the majority of children with appropriate follow-up. a significant number of children with EFI were not evaluated for a potentially treatable cause. Differences in approaches have been the use of flexible endoscopes by gastroenterologists and the primary use of rigid scopes by surgeons. Recent collaborative work with our surgery department has increased the emphasis on our findings. hot pocket. The mucosa was visually normal in all of the children in a previous study and in only 18% of the children in the present study (7). and mucosal biopsies are increasingly being procured. DISCUSSION In the present study. Number 1. In contrast. previous studies in adults have demonstrated an underlying pathology in the majority of subjects (2. treatable cause for EFI. food was the impacted object in only 10%. The small number of patients limited the ability to test for statistical significance. gross evidence of inflammation. and limited follow-up identified EoE and fixed strictures (2 s/p TEF/EA repair) as the underlying etiologies in 9 of the 12 children with repeat endoscopy. not chewing adequately. white exudates. The racial diversity of our patients studied is reflective of our local population. and shrimp in addition to ‘‘meat’’ as the identified substance. esophageal food impaction has frequently been assigned to chance events related to eating too quickly. and family history of esophageal strictures.org TABLE 3. this procedure is shared between gastroenterology and surgery services. and elevated eosinophils on mucosal sampling (>20/hpf) in 19 children.jpgn. Our data show an identifiable etiology in 92% of children with EFI in whom follow-up was available. including 2 following TEF/EA repair. . Despite the normal gross mucosal appearance in previous studies. In our study.4). The underlying cause of EFI in children has been uncertain. had gross evidence of esophagitis and abnormal histology. and ‘‘food’’ as the identified substance. Of the 4 with stricture at follow-up endoscopy. January 2011 Etiology of Esophageal Food Impactions in Children TABLE 2. Alternative causes such as esophagitis and strictures were present in 21 children. In our institution.JPGN  Volume 52. Unauthorized reproduction of this article is prohibited. In this regard. such as male predominance. Inflammation and stricture were present in 41 (85%) of these children. allergic history. y Other foods included burrito. 2 had anastomotic strictures following TEF/ EA repair and 2 had no previous esophageal surgery history but had persistent stricture for both first and second endoscopy. trachealization (circular rings). the majority of our patients FIGURE 2. due to our study design we are unable to fully quantify the percentage with EoE and GERD. Food impaction substance Subjects requiring endoscopy n ¼ 49 (%) Meat Chicken Beef Hot dog Other Vegetable/fruit Othery 41 11 14 7 9 3 5 (84) (22) (30) (14) (18) (6) (10) Spontaneous resolution n ¼ 16 (%) 10 5 3 2 (63) (31) (19) (13) 0 2 (13) 4 (25)  Other meats included lamb. Less than 50% of the children had biopsies taken at the time of disimpaction. Our study contributes to a growing body of literature that EoE represents an alternative.8). it is likely that a significant portion also have an element of GERD contributing to their inflammation. white plaques. the pathology demonstrated esophagitis in the majority of subjects (7. www. all of the clinical evidence documented that EFI occurred secondary to EoE in 5 of the 12 children and esophageal strictures in another 4. Often children who present acutely to an emergency room with EFI receive evaluation and disimpaction. Important to our study is the fact that no mucosal biopsies were obtained in any procedure performed by surgeons. The same studies documented between 83% and 100% of people with EFI reported prior episodes of EFI compared with 45% in our study (7. but follow-up is infrequent. whereas 4 showed persistent strictures and 3 were grossly normal. Esophageal eosinophilia is an increasingly common finding. we identified that of all of the children presenting acutely with esophageal foreign body impaction. As such. Endoscopic disimpaction was safe and effective in all children. with diverse underlying etiologies including GERD and EoE (11– 13). Taken together. Summary of endoscopic evaluations by pediatric gastroenterologist and surgeon (n U 49 (%)) Normal Abnormal Esophagitis Longitudinal furrows Inflammation Trachealization (circular rings) White plaques Nodularity Stricture s/p TEF/EA repair 9 (18) 40 (82) 27 (55) 12 (25) 10 (20) 8 (16) 7 (14) 4 (8) 13 (27) 5/13 TEF/EA ¼ tracheoesophageal fistula/esophageal atresia. Although the evidence suggests EoE in a large portion of the study population. sausage. A paucity of studies have measured the etiology of EFI in children.8). the majority of children with EFI have an underlying and often treatable cause. or postsurgical remodeling. In addition. Annual number of esophageal food impactions seen in the emergency department at The Children’s Hospital in Denver. Thus. Our study is limited in its retrospective nature.8). a number of features provide supportive evidence for EoE as the cause in a significant portion of children with EFI. bison.

Furuta GT.19:402–7.  3 subjects with biopsy only identified as esophagus.jpgn. REFERENCES 1. this was not the focus of the present study. 2009. Straumann A. 24:173–82. http://statpages. HPF ¼ high-power field (40). we recommend that all children with EFI should have mucosal biopsies obtained at the time of endoscopic disimpaction. 10. Aliment Pharmacol Ther 2006. Gastrointest Endosc 2005.37: 498–503. J Pediatr Surg 2001. 8. In addition. Am J Gastroenterol 2008. leaving their ultimate diagnosis uncertain. EFI difficult to evaluate.53:193–8. Esophageal food impaction: epidemiology and therapy.52:717–21. Review article: the pathogenesis and management of eosinophilic oesophagitis. 5. Rodrigo S. A retrospective. et al. Dig Dis Sci 2007.12:2328–34. J Pediatr Gastroenterol Nutr 2003. et al. 4. Hernandez-Peredo G. Esophageal eosinophilia in children with dysphagia. Biopsy pattern and histological analysis of mucosal inflammation (H&E stain) (n U 24)  Biopsy pattern Eos > 20/HPF (n ¼ 15) Eos 1–20/HPF (n ¼ 3) Proximal esophagus only Distal esophagus only Proximal and distal esophagus 0 3 2 with degranulation 0 1 3 2 with degranulation 1 20 9 7 with degranulation 2 Eos ¼ eosinophils. not distal or proximal. Because our results reveal that the majority of children have an underlying potentially treatable cause of EFI. Gastroenterology 2007. Desai TK. Number 1. . 2009. Retrospective analysis of esophageal food impaction: differences in etiology by age and gender. observational study. Although it would be interesting to compare features of EoE in patients who present with EFI with those who do not develop EFI. Stecevic V. Accessed September 1.36:1397–400. et al. www. et al. Eosinophilic esophagitis in children and adults: a systematic review and consensus recommendations for diagnosis and treatment. Pediatr Emerg Care 2003. Longstreth GF. Accessed September 1. et al. 13. Unauthorized reproduction of this article is prohibited. Collins MH. 11. 6 subjects had normal mucosa. et al.133:1342–63. 46 3. Oh D. 9. J Clin Gastroenterol 2008. 2.103:435–42. World J Gastroenterol 2006.org. Gastrointest Endosc 2001. Our retrospective study shows that food accounts for 10% of all esophageal foreign bodies in children presenting to a tertiary care children’s emergency department. Panagiotakis PH. Liacouras CA. Oliver MR. Longstreth KJ. Eosinophilic esophagitis: a newly established cause of dysphagia. Association of eosinophilic inflammation with esophageal food impaction in adults. Bostwick HE. Shaffer EA. Cheung KM. Predictors of recurrent esophageal food impaction: a case-control study.org Copyright 2010 by ESPGHAN and NASPGHAN. Furuta GT. 7. Acute food bolus impaction without stricture in children with gastroesophageal reflux. Boyd-Enders FT. Abboud G. Yan BM. 6. www. we recommend appropriate follow-up be scheduled following the acute event to allow for diagnosis and management of the underlying etiology. January 2011 TABLE 4.com. Reddy JG. Byrne KR. et al. High intraepithelial eosinophil counts in esophageal squamous epithelium are not specific for eosinophilic esophagitis in adults. Hilden K. Chang CH. Esophageal food impaction in children. The small percentage of children who presented for follow-up makes conclusions regarding etiology of EFI incomplete.Hurtado et al JPGN  Volume 52. Prasad GA. Molina M. 12.graphpad.42:771–5. Vicente Y. Yao JF. <50% received mucosal sampling. Cameron DJ. et al. Berezin S. Lao J. whether that is treatment for GERD and/or EoE.61:795–801. Although 75% required disimpaction.