Flexible endoscopy in children was first performed in the early 1970s when fiberoptic bronchoscopes were used for pediatrie gastrointestinal endoscopy.1 In the intervening years, diagnostic and therapeutic endoscopy have become routine in the evaluation of children with gastrointestinal and liver diseases. Advances in our understanding of conscious sedation and cardiovascular monitoring in infants and children have allowed these procedures to be performed safely.2
Many of the indications for pediatrie endoscopy are straightforward. There are, however, situations where the timing of endoscopy is unclear, such as in children with recurrent abdominal pain. In an effort to provide more concrete guidelines, the North American Society for Pediatrie Gastroenterology and Nutrition recently published a position paper outlining the indications for pediatrie gastrointestinal endoscopy.3 Notwithstanding these recommendations, the question of the utility of routine endoscopie examination in children with recurrent abdominal pain remains unanswered. Further well-controlled studies need to be conducted to resolve this issue. This article will review the various endoscopie procedures, their indications, and their diagnostic and therapeutic applications (Table).
A wide variety of flexible endoscopes are available for use in the pediatrie patient. The small diameter of new endoscopes (5.3 mm) allows for complete esophagogastroduodenoscopic examination in a newborn. This same endoscope also can be used as a flexible sigmoidoscope or colonoscope in newborns. Image resolution is excellent, and the biopsy channel is large enough to permit adequate biopsies for pathologic analysis.
Diagnostic esophagogastroduodenoscopy is the most frequently performed endoscopie procedure in children. Esophagogastroduodenoscopy is easily performed in the outpatient setting. The preoperative evaluation should include inquiry about risk factors for bacterial endocarditis and prophylaxis should be given when appropriate. The patient receives nothing by mouth for 6 to 12 hours before the procedure. Most pediatrie gastroenteroíogists directly administer intravenous sedatives in the endoscopy unit. A minority of pediatrie gastroenteroíogists advocate routine use of general anesthesia or intravenous anesthesia administered by an anesthesiologist. Most often, intravenous meperidine and midazolam are used in combination. Other intravenous agents such as fentanyl, ketamine, and propofol are used less often. An esophagogastroduodenoscopy takes between 5 and 20 minutes to complete, depending on the purpose for the examination. The patient's pulse, respiratory rate, blood pressure, and oxygen saturation are monitored continuously during the examination and during recovery. Complications are rare, but include oversedation, perforation, bleeding, aspiration, bowel wall hematoma, and broken teeth. The patient is almost always ready to resume usual activities the day following endoscopy. Indications for esophagogastroduodenoscopy that are likely to be encountered in a busy general pediatric practice are discussed below.
Epigastric Abdominal Pain, Acid-Peptic Disease, and Helicobacter pylori Infection
Reference should be made to the excellent discussion of the clinical evaluation of dyspepsia found elsewhere in this issue. In children with dyspeptic symptoms that are sufficiently significant to merit additional evaluation, esophagogastroduodenoscopy is the investigation of choice. Definitive diagnosis of esophagitis, gastritis, and peptic ulcer disease in children is made by visualization of the gross and microscopic findings. Endoscopie mucosal biopsies are easily obtained and should be performed in nearly all cases. The sensitivity and specificity of endoscopie examination appears to be far superior to barium radiography. For example, in one study of 89 children with an abnormal upper gastrointestinal series, 56 (63%) had normal upper endoscopies.4 Other studies5'6 have reported endoscopically detectable pathology in children with normal barium studies.
It is now well established that Helicobacter pylori causes gastritis and the majority of peptic ulcer disease in children and adults. Studies in pediatrie populations have determined that 15% of all children undergoing endoscopy and gastric biopsy have evidence of H pylori infection.7,8 Although serologie testing is available and a urea breath test will soon be available, endoscopie biopsy remains the standard for diagnosis. Biopsy specimens stained with hematoxylin-eosin or giemsa stains can demonstrate the presence of this helical organism in the mucus layer overlying the gastric mucosa. In addition, rapid methods for detecting urease, an enzyme found in high concentrations in Helicobacter, are available. This simple test requires placement of a gastric mucosal biopsy sample in a small pH-sensitive agar button. Hydrolysis of urea results in alkalinization of the agar and a readily visible color change in 2 to 24 hours.9 Some centers routinely culture biopsy material for H priori, but this method is not widely available.
Pediatric gastroenterologists have long realized that biopsy of the upper gastrointestinal mucosa is mandatory for a complete endoscopie assessment of the patient with upper gastrointestinal complaints. It is clear that significant microscopic abnormalities may exist in the context of a grossly normal endoscopie appearance. In some cases, it may be difficult to determine if gross and microscopic findings are primary or simply the result of symptoms such as persistent vomiting. An experienced pediatrie gastroenterologist will synthesize a final diagnosis and opinion only after thoughtful integration of the clinical evaluation, the gross endoscopie findings, and the mucosal histopathology.
Hematemesis of Melena
The etiology of upper gastrointestinal bleeding can be identified endoscopically in 85% to 90% of children.10,11 A negative endoscopie examination may imply a nasopharyngeal source for the bleeding. In addition to identifying the source of bleeding, prognostic information regarding the likelihood of rebleeding can be obtained.12,13 For example, an ulcer with a visible vessel has a high likelihood of rebleeding. Esophagogastroduodenoscopy also may provide the opportunity for therapeutic intervention. Sclerotherapy or banding of esophageal varices in patients with portal hypertension is used in the management of acute bleeding and to prevent further bleeding episodes.13 Nonvariceal hemorrhage can be managed by injection of sclerosing or vasoconstrictive agents and thermal coagulation using monopolar and multipolar probes, heater probe, or laser.14
Indications and Contraindications for Pediatric Gastrointestinal Endoscopic Procedures
Dysphagia and Odynophagia
Dysphagia is defined as difficulty swallowing and odynophagia as painful swallowing. These are relatively common complaints in a pediatrie gastroenterology practice. These symptoms are usually due to esophagitis as a result of gastroesophageal reflux.'3 Dysphagia in a patient with chest pain, "heartburn," recurrent vomiting, irritability, or sleep disturbances suggests the diagnostic possibility of esophagi tis. Esophageal pathology of sufficient severity to cause dysphagia merits endoscopie examination in most cases and empirical therapy is not advisable. Esophagogastroduodenoscopy allows for both gross and histologie evaluation of esophagitis and the detection of associated lesions such as peptic strictures and Barrett's esophagitis. Esophageal diverticula, esophageal infections, foreign bodies, vascular rings, and other extrinsic lesions also may cause dysphagia.
Esophageal strictures may result from peptic esophagitis or caustic ingestion, or may occur at an anastomotic site in children who have undergone surgical repair of esophageal atresia. Esophagogas.troduodenoscopy is both diagnostic and therapeutic as esophageal strictures can be dilated successfully with balloon catheters passed through the endo' scope. The success of bougie dilation may be monitored by follow-up endoscopy. Refractory strictures may be treated successfully with endoscopically directed intralesional steroid injections.16
Patients with a history of caustic ingestion require endoscopie examination of the esophagus and stomach within the first 24 to 48 hours following the ingestion to define the extent of the acute injury. Asymptomatic patients without evidence of oral or hypopharyngeal injury require examination as these criteria do not identify all patients with esophageal or gastric injury. It may be difficult to determine the potential of a given household product or chemical for "caustic" injury, and concrete clinical guidelines based on published evidence are not available for most agents. The most common product ingested is household bleach, and sufficient experience with ingestion of bleach by children has resulted in the conclusion that esophageal injury is unlikely. It is our practice to forego endoscopie examination in children who have ingested bleach unless there are unusual historical or clinical findings.
Drain cleaners, industrial cleaners, and crystalline detergents may cause significant injury and always require investigation by endoscopy. Consultation with the regional poison control center may be valuable in identifying the potential risk for chemical injury and the need for endoscopy. Disk (eg, calculator, hearing aid, and watch) batteries impacted in the esophagus require prompt endoscopie removal because of the risk for mucosal injury. Batteries that have reached the stomach can be managed conservatively unless they are large (>23 mm in diameter).
Endoscopie examination within the first 12 hours following the ingestion may not reveal the full extent of the injury, but sufficient information can be gained to permit a decision regarding the necessity for hospitalization and treatment. Barium studies are not reliable in detecting acute injury or in predicting stricture formation, but they are valuable 3 to 4 weeks after the ingestion to detect possible stricture formation.17
Foreign Body Ingestion
As a general rule, any object in the esophagus should be removed promptly regardless of symptoms. Coins account for the majority of esophageal foreign bodies in children. Most esophageal foreign bodies lodge at the level of the cricopharyngeal muscle, the level of the aortic arch, or at the gastroesophageal junction. The only possible exception to the rule of prompt removal is the coin lodged in the distal esophagus. A significant number of these pass spontaneously within 12 hours; if not, endoscopie removal is necessary. A review of radiographs from 80 children with a history of coin ingestion found that the coin had lodged in the esophagus in 25. Interestingly, 14% of these were asymptomatic.18 Coins had passed beyond the esophagus in 44 patients and no coin was identified in 11. A coin in the stomach should be managed conservatively as long as the patient is asymptomatic. All subsequent stools should be examined for the coin. If no coin is observed in the stool after 4 to 6 weeks and a follow-up radiograph shows the coin remains in the stomach, most gastroenterologists advise endoscopie removal.
Most foreign bodies in the stomach can be managed conservatively as described above. Even straight pins, nails, screws, and thumbtacks can be left in the stomach. Straight, sharp objects such as needles and long toothpicks (s*5 cm in older children and 3*3 cm in young children and toddlers) should be removed from the stomach endoscopically.
Unexplained vomiting may require endoscopie evaluation to exclude inflammatory disease (ie, Crohn's disease), infectious disease (ie, H pylori), or peptic ulcer disease. In addition, complications of vomiting such as esophagitis, Mallory-Weiss tears, and peptic strictures may be identified. Contrast radiologie techniques are not sufficiently sensitive to identify most patients with these diagnoses and complications. Barium upper gastrointestinal radiographs may complement information obtained by endoscopy in cases of unexplained vomiting by providing additional gross anatomical information and clues regarding upper gastrointestinal motility.
Evaluation of children with vomiting must include careful attention in the history and the physical examination to the signs and symptoms of the various "nongastrointestinal" causes of vomiting. These include neurological disorders and increased intracranial pressure, metabolic diseases, renal abnormalities, psychological disorders (anorexia nervosa and bulimia), toxic ingestion, and pregnancy.
Malabsorption should be suspected in children with weight loss, growth deceleration, or chronic diarrhea. These significant complaints usually present with variable signs and symptoms of abdominal distention, flatulence, excessive eructation, hyperphagia, anorexia, vomiting, abdominal pain, and edema. Screening tests, such as a random stool examination for fat or a 72-hour fecal fat, a serum d-xylose absorption test, stool-reducing sugars, stool a,-amitrypsin, and serum protein measurements provide further information regarding the possibility of small bowel mucosa! abnormality. Ultimately, a small bowel biopsy is required for diagnosis. The availability of fiberoptic endoscopes for infants and children as well as increased experience in the histopathologic interpretation of endoscopie biopsies have greatly diminished the use of the cumbersome technique of peroral capsule biopsy (Crosby capsule). An endoscopically obtained small bowel biopsy is simple and is almost always sufficient to allow an accurate diagnosis.19
Placement of Enterai Feeding Catheters
Percutaneous endoscopie gastrostomy tube placement was first described in 1980.20 Percutaneous endoscopie gastrostomy tube placement has become one of the most common therapeutic endoscopie procedures performed by pediatrie gastroenterologists. The primary indications are inability to swallow or an inability to maintain adequate oral intake. The technique is relatively simple and can be done without general anesthesia in many instances. The endoscope is used to insufflate the stomach, and the bright fiberoptic light is used to transilluminate the proper insertion site on the anterior abdominal wall. A small incision is made at this site on the abdominal wall and a 14- to 16-ga catheter is used as a trocar to enter the stomach. The endoscope is used to retrieve a guidewire placed through the trocar, and the percutaneous endoscopìe gastrostomy tube is passed orally over the wire and pulled through the abdominal incision. The endoscope is passed a second time to confirm proper positioning of the percutaneous endoscopie gastrostomy tube. A variety of complications have been reported including cellulitis, gastrocolic fistula, fasciitis, peritonitis, and exacerbation of gastroesophageal reflux. Many of these complications can be circumvented by exercising added caution in patients who have undergone previous laparotomy. Children with multiorgan dysfunction, especially cardiopulmonary dysfunction, should not undergo percutaneous endoscopie gastrostomy tube placement.
Placement of a percutaneous endoscopie gastrostomy tube in a patient with gastroesophageal reflux requires special consideration. Gauderer21 found that 12.9% (25/194) of patients (all neurologically impaired) with percutaneous endoscopie gastrostomy tubes subsequently required antireflux procedures. Therefore, some physicians recommend that neurologically impaired children and others with an increased risk of reflux have placement of a traditional open gastrostomy tube along with antireflux surgery. Others suggest 24-hour pH probe testing to evaluate the degree of reflux prior to percutaneous endoscopie gastrostomy tube placement. ll It may be helpful to treat reflux pre-percutaneous endoscopie gastrostomy tube with medications to identify those patients who respond to therapy.
Other Indications for EsophagogastroduodenoscopY
Other indications for esophagogastroduodenoscopy may be encountered. Occasionally, pediatrie gastroenterologists are asked to perform an endoscopie examination to evaluate the significance of abnormal findings on a contrast upper gastrointestinal series. In our experience, the most common request is for evaluation of a possible antrat (prepyloric) web in a child with chronic vomiting. Surveillance of the upper gastrointestinal tract in patients with a confirmed diagnosis of a familial polyposis syndrome is also indicated.
Colonoscopy and flexible sigmoidoscopy also are performed as outpatient procedures. A variety of techniques for bowel preparation can be used. The type of preparation depends in part on the age of the patient, the underlying condition, and whether pancolonoscopy is required. Most protocols require a clear liquid diet for 12 to 72 hours, as well as oral laxatives and enemas prior to the procedure. Polyethylene glycol solutions can be used as an osmotic purgative in older children and teenagers. Regardless of the regimen, adequate preparation is an essential precursor to a safe, informative examination of the large bowel.
Occasionally, a limited flexible sigmoidoscopy can be done without sedation. Otherwise, procedures for sedation and cardiovascular monitoring are similar to those described for upper endoscopy. A limited flexible sigmoidoscopy can be completed in 10 to 15 minutes while a pancoìonoscopy with polyp removal and retrieval make take as long as 2 hours. Complications of colonoscopy in children are rare, but include dehydration and electrolyte imbalance, oversedation, perforation, bleeding, and bowel explosion. Again, prophylaxis for bacterial endocarditis should be undertaken in patients at risk. Antibiotic prophylaxis for immunodeficient patients and in patients with indwelling catheters should be considered on an individual basis.
The indications for colonoscopy and flexible sigmoidoscopy are more straightforward than those for upper endoscopy. Lower gastrointestinal bleeding is the most common indication for endoscopie evaluation of the large bowel. In most instances, the pediatrie gastroenterologist will prepare the patient for colonoscopy, although a flexible sigmoidoscopy may be all that is required. As the procedure is being done, the extent of evaluation required to arrive at the proper diagnosis is determined by the endoscopist. A colonoscopy provides the gastroenterologist with the ability to visualize anatomy, make a diagnosis based on gross findings, obtain tissue to confirm the diagnosis, and possibly provide treatment to certain focal lesions. Note that abdominal pain is rarely an indication for colonoscopy in children. In the absence of the associated signs and symptoms noted below, there are no colonie lesions that present primarily with pain.
Children with colitis may have rectal bleeding, tenesmus, crampy abdominal pain, diarrhea, mucoid or mucopurulent stools, fever, growth deceleration, chronic anemia, and hypoproteinemia. A history of antibiotic exposure should be obtained. In most patients with acute symptoms, stool cultures for bacterial pathogens (including Escherichia coli 0157:H7) should be completed prior to colonoscopy. In geographic areas where Entamoeba histolytica is endemic, stool should be tested for this pathogen as well. If the investigation for infectious colitis is negative and the aforementioned symptoms continue, then endoscopie examination is merited. However, it should be recognized that there may be substantial overlap between both gross and microscopic findings in early inflammatory bowel disease and persistent bacterial colitis. Therefore, caution should be exercised in conveying the diagnosis of inflammatory bowel disease as "final" in these uncertain cases.
Colonoscopy readily demonstrates colitis. Loss of the normal submucosal vascular pattern, excessive friability, bleeding, and ulcération indicating colitis are all suggestive gross endoscopie findings. Other findings include pseudomembranes, which have a highly characteristic appearance and are nearly pathognomonic for pseudomembranous enterocolitis due to Clostridium aiff\d\e toxin. Also, the distribution of mucosal abnormalities may have diagnostic significance in patients with inflammatory bowel disease.
For example, continuous disease suggests ulcerative colitis while "skip lesions" suggest Crohn's disease. The terminal ileum can be reached in the majority of children and the presence of disease in this location confirms the diagnosis of Crohn's disease. Lymphoid nodules in the terminal ileum may be quite prominent in children, and on occasion may be mistaken for inflammatory disease, especially when the endoscopist is not experienced with pediatrie ileoscopy. Biopsy specimens may confirm the type of colitis, but "indeterminant" cases are relatively common. In patients with inflammatory bowel disease, all the clinical and endoscopie variables should be integrated to make a diagnosis because assignment as Crohn's disease or ulcerative colitis may affect short- and long-term medical and surgical management.
Juvenile polyps are a common cause of rectal bleeding in children, especially early in the first decade of Ufe. Bleeding due to polyps is usually painless, intermittent, bright red, and streaked onto the stool. Colonoscopy is indicated in children with these findings. Polyps are removed by electrocautery. At least 25% of children have more than one polyp, and pancolonoscopic examination is indicated. It is important to retrieve as many polyps as possible for histopathology. Histopathologically, most are hamartomatous polyps with a heavy infiltration of "inflammatory" cells, although adenomatous changes are increasingly recognized in juvenile polyps. Most pediatrie gastroenterologists recommend follow-up examination in children who have polyps with adenomatous features, although extensive further study will be required to define the malignant potential. Children in families with familial adenomatous polyposis coli should be screened by colonoscopy on an annual or biannual basis beginning at the age of 5.
Vascular lesions (ie, Rendu-Osier- Weber and blue rubber bleb nevus syndrome) are unusual causes of rectal bleeding in children. Techniques of endoscopie hemostasis such as epinephrine injections, heater probe and multipolar electrocoagulation, rubber band ligation, and laser therapy have been successfully and safely applied in children.23
Cancer Surveillance and Prevention
Annual or biannual pancolonoscopy is recommended for cancer surveillance in patients with ulcerative colitis for 7 to 10 years or longer. Multiple biopsies are obtained in a serial fashion to evaluate for the presence of dysplasia, which is considered a preneoplastic lesion.24'25 Surveillance also is recommended in patients with familial adenomatous polyposis, as mentioned above.
CHALLENGES FOR THE FUTURE
There are many challenges for the future of pediatrie endoscopy. Foremost is the design of wellcontrolled analyses of the indications for endoscopy and outcome studies of children managed with and without endoscopie examination. New pharmacological agents for conscious sedation should be studied in a formal manner. Further technological innovation in the design of endoscopes for small children, especially premature neonates, is needed. Specialty instruments such as side-viewing scopes for endoscopie retrograde cholangiopancreatography need further development. Therapeutic endoscopy in pediatrie patients is still in the early stages, and major new advances are certain to occur.
1. Freeman NV. Clinical e valuation of the fiberoptic bronchoscope (Olympus BF 5B) for pediatric endoscopy. J Pedían· Stag. 1973;8:213-220.
2. Ament ME, Berquist WE, Vargas J, et al. Fiberoptic upper intestinal endoscopy in infants and children. Pediatr Clin North Am. 1988;35:141-155.
3. Squires RH, Colletti RB. Indications far pediatrie gastrointestinal endoscopy: a medical position statement of the Notth American Society for Pediatric Gastroenterology and Nutrition. J Pediatr Gastroenterol Nutr 1996;23:107-110.
4. Miller V, Doig CM. Upper gastrointestinal tract endoscopy. Arch Dis Chili 1984;59:1100-1102.
5. Tedescp FJ, Goldstein DD, Gleason WA, et al. Upper gastrointestinal endoscopy in the pediatric patient. Gastroenterohgy. 1976;70:492-494.
6. Gleason WA, Tedesco FJ, Keating JF, et al. Fiberoptic gastrointestinal endoscopy in infants and children. J Pediatr. 1974;85:810-813.
7. Hassall E, Dimmidk JE. Unique features of Helicobacter pylori disease in children. Dig Dis Sci. 1991;316:417-423.
8. Kilbridge PM, Dahms BB, Czinn J. Campylabacter pylori-associated gastritis and peptic ulcer disease in children. Am J Dis Child. 1988;142;1149-1152.
9. Czinn SJ, Carr H. Rapid diagnosis of Campylobacter pylori-associated gastritis. J Pediatr. 1987;110:569-570.
10. Cadranel S, Rodesch P, Peters J, et al. Fiberendoscopy of the gastrointestinal tract in children. A series of 100 examinations. Am J Dis Child. 1977;131:41-45.
11. Frolla J, Diehl AS, Benvenuti GA, et al. Upper gastrointestinal fiberoptic endoscopy in pediatric patients. Gastrointest Endosc. 1983;29:Z79-281.
12. Sutton F. Upper gastrointestinal bleeding in patients with esophageal varices. What is the most common source? Am J Med 1987;83:273-275.
13. Thapa BR, Mehta S. Endoscopic sclerotherapy of esophageal varies in infants and children. J Pediatr Gastroenterol Nutr. 1990; 10:430-434.
14. Chung SC and others. Injection or heat probe for bleeding ulcer. Gastroenterology. 1991;100:33-37.
15. Black DD, Haggitt RC, Orentein SR, et al. Esophagitis in infants: morphometric histologic diagnosis and correlation with measures of gastroesophageal reflux. Gastronterology. 1990;98: 1408-1414.
16. Berenson GA and others, Inrralesional steroids in the treatment of refractory esophageal strictures. J Pediatrr Gastroenterol Nutr. 1994; 18:250-252.
17. Boria AR, Ransdell HT Jr. Thomas TV, et al. Lye injuries of the esophagus: analysis of 90 cases of lye ingestion. J Thorac Cardiovasc Sugf. 1969;57:533-538.
18. Hodge D, Tecklenburg F, Fleisher G. Coin ingestion: does every child need a radiograph! Ann Emerg Med 1985; 14:443-446.
19. Variend S, Phillips Au, Walder-Smith JA. The small intestinal mucosal biopsy in childhood. Perspec Pediatr Pathol 1984;1;57-78.
20. Cauderer MWL, Ponsky JL, Izant RJ Jr. Gastrostomy without laparocomy: a percutaneous endoscopic technique. J Pediatr Surg. 1980; 15:872-875.
21. Gauderer MWL. Percutaneous endoscopic gastrostomy: a 10-year experience with 220 children. J Poto Siog. 1991;26:301-305.
22. Wheatley MJ and others. Long-term follow-up of brain-damaged children requiring feeding gastrostomy: should an antireflux procedure always be performed? J Pediatr Surg. 1991;26;301-305.
23. Noronha P, Leist M. Endoscopic laser therapy for gastrointestinal bleeding from congenital vascular lesions. J Pediatr Gastroenterol Nutr. 1988;7:375-378.
24. Lofberg R and others. Colonoscopic surveillance in long-standing total ulcerative colitis-a 15 year follow-up study. Gastroenterology. 1990;99:1021-1033.
25. Woolrich AJ, DaSilva MD, Korelitz BI. Surveillance in the routine management of ulcerative colitis: the predictive value of low-grade dysplasia. Gasaoenteroiogy. 1992;103:431-438.
Indications and Contraindications for Pediatric Gastrointestinal Endoscopic Procedures