The threshold to perform paediatric endoscopy, especially in infants, has changed recently, with better anaesthetic techniques and technological advances in the size and flexibility of specially designed paediatric endoscopes (1,2). These changes, combined with increasing indications, have resulted in an increased number of endoscopies performed in infancy. Despite this, data on gastrointestinal mucosal histological findings in infants are limited to occasional small and selected series only (3–5). Presently, there are no systematic data regarding the usefulness of endoscopy in the first year of life.
The aim of the present study was to identify infants younger than 1 year from a single tertiary paediatric gastroenterology unit who underwent gastrointestinal endoscopy and review clinical indications and histological outcomes.
Patients referred to our paediatric gastroenterology centre during the period June 1987 to August 2007 who underwent gastrointestinal endoscopy during the first year of life were retrieved from hospital databases. From the initial 1319 cases fulfilling the above criteria, 295 cases in which there was insufficient clinical information or the gastrointestinal tract was not intact (eg, short gut, Hirschsprung disease, intestinal atresia) were excluded from the study. Clinical indications and histological findings were therefore reviewed in 1024 cases (median age 210 days, range 6–365 days) in a total of 823 infants (433 boys, 390 girls).
Nine hundred thirty-three gastroesophageal duodenoscopies and 439 colonoscopies were performed during the study period. In order of frequency, the main clinical indications were diarrhoea (522/1024, 51%), failure to thrive (428/1024, 41.2%), symptoms of reflux (278/1024, 27.1%), and rectal bleeding (87/1024, 8.5%), as shown in Figure 1.
Reflux symptoms were the indication for endoscopy more frequently in infants ages 6 to 12 months compared with the first 6 months of life (30.6% vs 22%, respectively, P < 0.01), whereas diarrhoea was a less common indication in the second half of the first year of life (60.3% vs 44.7%, P < 0.001), but representing the predominant symptom warranting endoscopic investigations across the entire age group (Fig. 2). There was no significant difference in the likelihood of faltering growth (Fig. 2) or bleeding per rectum (data not shown) as an indication for gastrointestinal endoscopy in the first or second half of the first year.
The procedure resulted in insufficient material for adequate histological assessment in only 25 cases (2%). Mucosal histological findings were normal in 346 of 1024 (33.8%) cases, whereas histological abnormalities were identified in 653 of 1024 (63.8%) cases (Fig. 3).
Histological abnormalities included oesophagitis in 121 of 431 (28%), gastritis in 92 of 431 (21.3%), and a range of enteropathic features in 437 of 588 (74.3%) cases, mainly in the form of villous atrophy and inflammatory cell infiltration. Colonic histology was abnormal in 234 of 325 cases (72%), including a range of inflammatory changes from mild nonspecific inflammation, eosinophilic infiltration, and active colitis (Fig. 4).
We subsequently analysed the histological findings in 348 patients who underwent both upper and lower gastrointestinal endoscopy at the same session. Eighty-five patients (24.4%) had normal histology throughout, 3 (0.9%) had inconclusive results from inadequate samples, whereas the remaining 260 (74.7%) had histological abnormalities in at least 1 site. Evidence of oesophagitis or gastritis was found in 26 of 260 cases (10%) and 29 of 260 cases (11.1%), respectively. Interestingly, isolated enteropathy was present in 74 of 260 cases (28.5%), abnormal colonic mucosa only was found in 59 of 260 cases (22.7%), whereas in the majority of the cases, both small bowel and colon were affected (107/260 cases, 41.1%). In the remaining 20 cases in which neither the small nor the large bowel was affected, inflammation was confined to the stomach and/or oesophagus. For the group of patients with abnormal histology in both small bowel and colon, main clinical indications in order of frequency were diarrhoea (66/107, 61.7%), failure to thrive (42/107, 39.2%), symptoms of reflux (23/107, 21.5%), and per rectum bleeding (16/107, 14.9%). Further analysis within this group demonstrated that in 28 of 107 cases, colitic involvement did not correlate with clinical symptoms such as bloody diarrhoea, constipation, or recurrent abdominal pain. In 10 of 107 cases, enteropathy was an unexpected finding because the clinical history was unhelpful with no diarrhoea, hypoalbuminemia, regurgitation, or failure to thrive.
More specific diagnoses made in this series included microvillous inclusion disease in 21 of 1024 (2%), autoimmune enteropathy in 12 of 1024 (1.2%), graft-versus-host disease post–haematopoietic stem cell transplantation in 9 of 1024 (0.9%), tufting enteropathy in 5 of 1024 (0.5%), and disaccharidase deficiency in 2 of 1024 (0.2%) cases (Fig. 5).
Relevant clinical information was available in 889 of 1024 episodes regarding postprocedure course, which revealed only 1 colonic perforation (1/889, 0.1%); this was in a neonate who underwent endoscopic examination at the age of 26 days, with severe inflammation and ulceration at the level of the sigmoid colon and underwent sigmoid resection the following day. The child was subsequently diagnosed as having intractable ulcerating enterocolitis of infancy. In this condition, the inflammatory process and ulceration may be transmural and covered by a thin layer of serosa only; hence, the risk of perforation in these children is greatly increased because of the nature of the disease process.
This is the first large study reporting on gastrointestinal endoscopy indications, histological findings, and outcomes in the first year of life. The histological findings when positive changed clinical management in a group in which both symptoms and signs are notoriously difficult to interpret. The findings of the present study demonstrate that gastrointestinal endoscopy with mucosal biopsies in this age group is a safe and greatly informative test, revealing abnormalities in approximately two-thirds of patients, with few samples uninterpretable.
Diarrhoea is the most common clinical indication for infants younger than 1 year to undergo gastrointestinal endoscopic investigation, followed by faltering growth, symptoms of reflux, and bleeding per rectum, with other symptoms being less common as presenting symptoms. In the majority of cases in which upper and lower endoscopy were performed during the same session, some degree of mucosal inflammation was evident in both small and large bowel, even in the absence of signs or symptoms directly related to the site of gut mucosal findings showing the importance of panendoscopy in this age group, because there is a high occurrence of occult mucosal histological findings contributing to the clinical picture in patients in whom clinical symptoms can be vague or nonspecific. Diarrhoea was a more frequent indication during the first 6 months compared with older infants, although it was the most common indication across the entire age group, whereas reflux symptoms were more likely to be reported in infants who underwent endoscopy between 6 and 12 months of age rather than in the first 6 months.
The importance of the histological assessment of endoscopic mucosal biopsies in determining diagnosis, management, and prognosis in infants with chronic diarrhoea is recognized increasingly. It is important to stress that this refers to infants with protracted diarrhoea in whom infectious causes have been excluded and initial dietary manipulations failed to lead to clinical improvement or resolution of the symptoms. Abnormal histological findings included enteropathic features and a range of inflammatory changes in the colon, such as mild nonspecific inflammation, eosinophilic infiltration, and active colitis. The spectrum of differential diagnosis of diarrhoea in this age group is broad, including conditions with a normal villus–crypt ratio, such as congenital defects in transport of electrolytes, glucose/galactose or congenital enterokinase deficiency, as well as entities manifested with villous atrophy, such as allergic enteropathy, infectious/postinfectious enteropathy, and other much rarer and inherited enteropathies (eg, microvillous inclusion disease, IPEX syndrome, tufting, and autoimmune enteropathy) (6,7). Early recognition of such disorders may prevent unnecessary drug therapy or adverse nutritional outcomes. Treatment options vary from dietary manipulations in food allergies to immunosuppressive treatment in autoimmune enteropathy or small bowel transplantation in microvillous atrophy and tufting enteropathy (8–10). Clinical conditions with pathognomonic histological features unfortunately reflect the minority of the cases. For the rest, findings are interpreted in the clinical context, and empirical treatment is used on many occasions. It is worth noting that in our study, autoimmune enteropathy was diagnosed in 10 patients, microvillous inclusion disease in 18, and tufting enteropathy in 4. This is probably a reflection of the referral pattern of a specialised tertiary paediatric gastroenterology centre.
Regurgitation was the third most common presenting complaint in our studied population. In 245 of 278 cases with symptoms of reflux who received gastroesophageal duodenoscopies, only 62 cases had histological evidence of inflammation in the lower oesophagus; however, of the 121 cases with histological features of oesophagitis identified from our total population, no specific related symptoms (regurgitation, haematemesis, irritability, feeding difficulties, faltering growth) were reported in 24 cases (data not shown). Poor correlation between symptoms suggestive of gastroesophageal reflux disease and histological oesophagitis has been well recognized in previous studies. In the study by Salvatore et al (11), oesophagitis was present in 17 of 44 infants with presumed gastroesophageal reflux disease, and there was no correlation found among clinical history, pH study, and oesophageal histology. Parental perception and report of symptoms, as well as visceral hypersensitivity, were among the possible explanations for this discrepancy.
Bleeding per rectum was a common reason to perform lower gastrointestinal endoscopy in this age group. In the present study, 47 of 87 (54%) patients presenting with rectal bleeding had histological evidence of colonic inflammation, with the suggestion of inflammatory bowel disease in 3 infants, whereas 19 of 87 patients had normal large bowel mucosal biopsies. Although cow's-milk protein allergy is a common cause in infants, other reasons for the bleeding may be present (12–14). Usually, dairy elimination would be the first approach, with endoscopic investigations reserved for the subgroup of patients with persistent symptoms or more severe presentation. In a study of 22 infants younger than 6 months presenting with rectal bleeding who underwent limited flexible sigmoidoscopy, allergic colitis was found in 14 patients, nonspecific colitis in 3, and 5 patients had normal mucosal biopsies (15). Rectal bleeding in all of the infants with normal biopsies or nonspecific mild inflammatory changes resolved without dietary change, with the exception of 1 infant with family history of ulcerative colitis who was subsequently diagnosed as having infantile inflammatory bowel disease. The authors commented that in more than one-third of cases, diet would have been unnecessarily changed. Additionally, Cannioto et al (16) drew attention to the fact that routinely treating patients for presumed cow's-milk protein allergy can cause diagnostic delays. Presence of eosinophils in colonic mucosa can be misleading and may represent early inflammatory bowel disease during the first 2 years of life, stressing the importance of early further investigations when elimination diet fails.
Selection and interpretation of the clinical information did not prove to be an easy task, such as in cases in which presumed diagnoses rather than clinical symptoms and signs were given or reported complaints were biased by physician's or parental perception. This probably explains symptoms such as abdominal pain in infants during the first year of life appearing among the clinical indications for endoscopy in our retrospective analysis. It stresses once more the importance of clinicians providing clear, accurate, and relevant clinical information to optimise the diagnostic yield of the requested tests.
Safety of gastrointestinal endoscopy performed in our centre in infants younger than 12 months was highlighted by the fact that there was only 1 incidence of colonic perforation in this 20-year period in a baby with severe intractable ulcerating inflammation. This is transmural disease with often only a serosal covering of the ulcers and would be categorized as a “high risk” endoscopy with an increased perforation rate. Endoscopic perforation of the colon in adults ranges from 0.03% to 0.15% in various studies (17,18). The largest single-centre paediatric study (19) reported colonic perforation rate of 0.09% in a total of 3269 colonoscopies in paediatric patients ages 0 to 18 years during a 26-year period. Two of the patients, ages 13 and 17 years, had known ulcerative colitis, whereas the third patient was a 6-month-old infant investigated for protein-losing enteropathy. Because no demographic data of the studied population were provided, the proportion of infants undergoing endoscopy is not known, and therefore direct comparison is difficult. The same study highlighted that upper gastrointestinal endoscopies can also result in 0.06% of cases in iatrogenic injury, defined as bleeding, perforation, or mucosal tear. Understandably, experienced operators, anaesthetic technique, and drugs combined with continuing technological developments in the area of endoscopy, with smaller and more flexible instruments, ensure that procedures are carried out in the safest way possible.
Our retrospective study spans a long period, and it is therefore expected that referral pattern, technical means, diagnostic workup, and treatment modalities have been modified. It also clearly illustrates the fact that gastrointestinal endoscopy is a greatly effective, informative, and clinically useful procedure in the first year of life, influencing management in the majority of the cases in which it was performed.
1. Amateau SK, Canto MI. Enhanced mucosal imaging. Curr Opin Gastroenterol 2010; 26:445–452.
2. Croffie JM. Advances and new technologies in adult endoscopy: can they be adapted to pediatrics? Curr Gastroenterol Rep 2007; 9:208–213.
3. Scudiere JR, Maitra A, Montgomery EA. Selected topics in the evaluation of pediatric gastrointestinal mucosal biopsies. Adv Anat Pathol 2009; 16:154–160.
4. Dupont C, Kalach N, de Boissieu D, et al. Digestive endoscopy in neonates. J Pediatr Gastroent Nutr 2005; 40:406–420.
5. Kirberg A, Latore JJ, Hartart ME. Endoscopic small intestinal biopsies in infants and children: its usefulness in diagnosis of celiac disease and other enteropathies. J Pediatr Gastroenterol Nutr 1989; 9:178–181.
6. Sherman P, Mitchell D, Cutz E. Neonatal enteropathies: defining the causes of protracted diarrhea of infancy. J Pediatr Gastroenterol Nutr 2004; 38:16–26.
7. Hartfield D, Turner J, Huynh H, et al. The role of histopathology in diagnosing protracted diarrhea of infancy. Fet Pediatr Pathol 2010; 29:144–157.
8. Thapar N, Shah N, Ramsay AD, et al. Long-term outcome of intractable ulcerating enterocolitis of infancy. J Pediatr Gastroenterol Nutr 2005; 40:582–588.
9. Ruemmele FM, Jan D, Lacaille F, et al. New perspectives for children with microvillous inclusion disease: early small bowel transplantation. Transplantation 2004; 77:1024–1028.
10. Paramesh AS, Fishbein T, Tschernia A, et al. Isolated small bowel transplantation for tufting enteropathy. J Pediatr Gastroenterol Nutr 2003; 36:138–140.
11. Salvatore S, Hauser B, Vandemaele K, et al. Gastroesophageal reflux disease in infants: how much is predictable with questionnaires, pH-metry, endoscopy and histology? J Pediatr Gastroenterol Nutr 2005; 40:210–215.
12. Machida HM, Catto Smith AG, Gall DG, et al. Allergic colitis in infancy: clinical and pathologic aspects. J Pediatr Gastroenterol Nutr 1994; 19:22–26.
13. Arvola T, Ruuska T, Keranen J, et al. Rectal bleeding in infancy: clinical, allergological, and microbiological examination. Pediatrics 2006; 117:e760–e768.
14. Murphy MS. Management of bloody diarrhoea in children in primary care. BMJ 2008; 336:1010–1015.
15. Xanthakos S, Schwimmer JB, Melin-Aldana H, et al. Prevalence and outcome of allergic colitis in healthy infants with rectal bleeding: a prospective cohort study. J Pediatr Gastroenterol Nutr 2005; 41:16–22.
16. Cannioto Z, Berti I, Martelossi S, et al. IBD and IBD mimicking enterocolitis in children younger than 2 years of age. Eur J Pediatr 2009; 168:149–155.
17. Lohsiriwat V, Sujarittanakarn S, Akaraviputh T, et al. What are the risk factors of colonic perforation? BMC Gastroenterol 2009; 9:71.
18. Cobb WS, Heniford BT, Sigmon LB, et al. Colonoscopic perforations: incidence, management and outcomes. Am Surg 2004; 70:750–758.
19. Iqbal CW, Askegard-Giesmann JR, Pham TH, et al. Pediatric endoscopic injuries: incidence, management, and outcomes. J Pediatr Surg 2008; 43:911.