Proctosigmoidoscopy and rectal biopsy are generally considered routine office procedures for the practicing internist and are regularly employed in the evaluation of patients with rectal bleeding, bloody diarrhea, chronic diarrhea, or abdominal pain. Yet the procedures are rarely used by pediatricians in investigating similar complaints in infants and small children. Adequate proctosigmoidoscopy is often replaced by more hazardous and less definitive radiographic procedures or inadequa te examination through a test tube in the rectum. Because a properly performed proctosigmoidoscopy is safe and often quite helpful in the diagnosis of acute and chronic rectosigmoid disease in infants and children, a discussion of this subject is an essential part of any review of modern-day pediatric gastroenterology.
Figure 1. Four sizes of sigmoidoscopes, measuring 25 cm. x 19 mm., 25 cm. ? 15 mm., 15 cm. ? 15 mm., and 10 cm. x 11 mm.
In this presentation, the proper technique for proctosigmoidoscopy in infants and children is described. The indications for doing the procedure and the diagnoses that can be made by proctosigmoidoscopy are discussed. The value of performing rectal biopsy is also cited, as it should usually be considered part of the complete proctosigmoidoscopic examination.
Before proctosigmoidoscopic examination is performed in any infant or preschool child, adequate sedation must be provided to avoid emotional trauma to the patient as well as to prevent patient movement, which may increase the risk of the procedure and can cause traumatic lesions in the mucosa.
All children less than seven years of age should be sedated with an intramuscular injection of meperidine (Demerol®), 2 mg./kg., promethazine (Phenergan®), 1 mg./kg., and chlorpromazine (Thorazine®), 1 mg./kg., 75 minutes before each examination, using one-half this amount in infants less than one month of age. A dose of 50 mg. of meperidine, 25 mg. of promethazine, and 25 mg. of chlorpromazine is not exceeded. Patients receiving this amount of sedation are usually asleep when the examination begins. The adequacy of sedation is tested before proctosigmoidoscopy by digital rectal examination. If the patient will permit such examination without moving about or straining, proctosigmoidoscopy can usually be performed without difficulty. If, however, the child moves about during digital examination, additional sedation can be provided with intravenously administered diazepam (Valium®). This agent must be given very slowly to avoid respiratory depression, and the dosage must be carefully titrated to provide just enough sedation for the examination to proceed. Children over seven years of age can usually undergo the procedure without sedation, but if they are anxious or uncooperative, they can be sedated with intravenously administered diazepam alone. In this setting, excessive dosage of the drug can produce a hyperactive state, making the patient less cooperative than before. Ten mg., given over a five-to-10-minute period, is usually an adequate amount in these patients.
Patients are placed in a kneeling position for the examination, and a standard sigmoidoscopy table is adjusted to the patient's size. Small infants are placed prone on the table, with a rolled sheet beneath the abdomen.
The examination begins with careful visual perianal inspection for fissures, fistulae, external hemorrhoids, or other abnormalities. Subsequently, digital rectal examination using water-soluble lubricant is performed to dilate and lubricate the anal canal as well as to assess the adequacy of sedation. In addition, the examiner palpates for the presence of any mass lesions or hard stool.
Figure 2. Proper position for infant sigmoidoscopy. The barrel of the instrument is held with the left hand, which rests on the patient's gluteal region.
The appropriate instrument is then selected for use. A variety of pediatric sigmoidoscopes are available in several sizes (Figure 1). We generally use a 10-cm.xll-mm. instrument for infants less than three months of age and a 15-cm.xl5-mm. instrument for children more than three months old but less than two years of age. Children over two years old can usually be examined with the standard 25cm.xl9-mm. adult instrument; however, narrower instruments must occasionally be used.
The sigmoidoscope is then lubricated with water-soluble lubricant and inserted with the right hand, the right thumb holding the obturator in place. The left hand is placed on the gluteal region throughout the examination, with the barrel of the instrument grasped between the left thumb and forefinger (Figure 2). This position permits the left hand and the instrument to move with the patient in case unexpected movement occurs. Holding the instrument with the right hand alone could result in the instrument's being forced against the mucosa, lacerating it or possibly perforating the colon. The instrument is advanced slowly through the anal canal with the obturator in place. As soon as the pressure required to insert the instrument decreases, the anal sphincter has been passed and the obturator can then be removed with the right hand. After this point in the examination, the instrument should be advanced only while the examiner is looking though it, and then only when lumen can be clearly visualized. As long as lumen is easily seen, the sigmoidoscope can be carefully advanced with safety.
Figure 3. Judicious use of air insufflation permits a more complete examination.
An air insufflation bag is attached to the instrument before the examination, as judicious insufflation of air often helps to find an obscure lumen (Figure 3). Excessive use of air produces patient discomfort, however, and should be avoided.
The examiner looks for the presence oí spontaneous friability, induced friability, and the presence or absence of ramifying mucosal blood vessels; these three observations have been shown to provide reliable interobserver agreement in adult patients with colitis.1,2 In addition, the presence of any discrete lesions and the extent and distribution of any inflammatory changes are noted. Although such findings as edema, wetness, pallor, granularity, and hyperemia are not generally considered reliable, they are frequently noted if they appear obvious. The presence of such findings alone, however, does not usually constitute an abnormal examination.
Induced friability is defined as bleeding of the mucosa after it has been wiped with a cotton swab. This is done by rotating a sigmoidoscopy swab 360 degrees against the mucosal surface. Removal of some of the cotton from the standard adult swabs with permit their passage through the smaller pediatric instruments. The mucosa should be wiped at least five times to exclude induced friability, which is often present in the absence of either spontaneous friability (bleeding of the mucosa before wiping) or nonvisualization of ramifying vessels. Failure to wipe the musosa can cause some examinations of patients with mild colitis to be considered normal.3 It is important to perform mucosal wipe tests as the instrument is being advanced, since confusion between scope trauma and true induced friability can result if the wiping is done when the instrument is withdrawn.
The presence of diarrheal stool does not prevent adequate examination, as it can often be suctioned out or wiped away with cotton swabs. Suctioning diarrheal stool is preferable to the use of cleansing enemas to remove it, as trauma induced from the enema apparatus or irritation from the cathartics used in enema preparations can often be confused with inflammatory changes in patients with mild disease. Unless a large amount of hard stool is present or unless one is planning to remove a polyp, patients should undergo sigmoidoscopy without prior cathartic or enema preparation.
Once the instrument can no longer be easily advanced, it is withdrawn in a circular manner. This permits complete re visualization of the entire mucosa and allows one to find any abnormalities that may have been missed during earlier phases of the examination. However, care must be taken to avoid misinterpreting instrument trauma as being due to diseased mucosa.
The indications for proctosigmoidoscopy in children are basically the same as those in adult patients, except that routine screening for malignancy is not necessary. Patients with bloody diarrhea, rectal bleeding, or chronic diarrhea of unknown cause should undergo proctosigmoidoscopy.3 Patients with chronic abdominal pain in whom the clinical course is compatible with inflammatory bowel disease should also undergo the procedure. This would include primarily patients with recurrent fever, weight loss, lack of appropriate growth velocity, arthritis, or other systemic complaints. In addition, patients with problems obviously related to rectal or perianal disease - such as painful bowel movements, excess mucus in the stool, perianal fissures or fistulas, or anal ulcers - also require proctosigmoidoscopy. Family members of patients found to have familial polyposis syndromes require proctosigmoidoscopy.
Proctosigmoidoscopy is valuable in patients with acute bloody diarrhea for a variety of reasons. Such patients usually have colitis, which is usually due to infection, drugs, Crohn's disease, or idiopathic ulcerative colitis.3 In infants' formula, proteins may cause colitis or entercolitis. Infectious causes of colitis include bacteria (salmonella, shigella, Escherichia coli, and staphylococcus), parasites (Entamoeba histolytica and Trichuris trichiura), and viral disease (lymphogranuloma venereum). Although findings at sigmoidoscopy are infrequently characteristic of a specific origin, certain factors may suggest the cause of colitis. Discrete ulcers are said to occur more frequently with amebiasis, although this finding is certainly not specific.4,5 Submucosal nodular appearance might suggest tuberculous enteritis, whereas a pseudomembrane could indicate shigellosis or staphylococcal enterocolitis.* Cultures for pathogenic bacteria, or wet mounts for amebae, may be better obtained by rectal swab than from stool sample alone.5
Drug-induced colitis is most often recognized as being due to clindamycin or lincomycin;7,8 however, other antibiotics - such as ampicillin, penicillin, and tetracycline - have also been implicated.9 This entity frequently, but not always, presents with a pseudomembrane, and the presence of a pseudomembrane in a patient who is taking antibiotics or has taken them within the previous six weeks is highly suggestive of this disorder.
Formula protein intolerance due to milk or soy protein10,1 ' can cause colitis during the first six months of life. This disorder may present with either rectal bleeding or bloody diarrhea. Because it has no characteristic appearance at proctosigmoidoscopy, it can be diagnosed only by withdrawal of the offending formula protein and rechallenge with the appropriate formula after all symptoms have ceased and mucosa has returned to normal.
Both idiopathic ulcerative colitis and Crohn's colitis may present during the first decade of life; however, Crohn's disease rarely appears before the second decade. These two diseases constitute the category of idiopathic inflammatory bowel disease and can be difficult to differentiate during the early stages.12 Rectal sparing or a cobblestone appearance of the mucosa is more likely to occur with Crohn's colitis. Colitis present only in the rectum - i.e., ulcerative proctitis - is a variant of ulcerative colitis characterized by a lower incidence of malignancy and mild, continuous symptoms despite therapy. It is therefore essential for the examiner to note the extent of mucosal disease as well as the character of any inflammatory changes.
Rectal bleeding in children may be the presenting symptom of colitis due to any of the above causes, but it may also result from fissures, polyps, discrete ulcers, traumatic lesions, ischemia, lymphoid hyperplasia, arteriovenous malformations, and congenital diverticula with gastric mucosa (i.e., Meckel's diverticulum). The source of bleeding can be determined by proctosigmoidoscopy in roughly three-fourths of patients with this complaint.3 This technique should be the first diagnostic procedure in such patients because it is safe and usually definitive.
Patients with chronic diarrhea may have abnormalities of mucosal function or motility of the small bowel, colon, or both. A logical approach to this problem certainly includes proctosigmoidoscopy as a means of ruling out the presence of colitis. Approximately one-third of patients with chronic diarrhea may have abnormal findings at proctosigmoidoscopy.3
In most cases, polyposis in members of families of patients with familial polyposis syndromes can be diagnosed early by proctosigmoidoscopy. Familial adenomatous polyposis of the colon is the most serious because of the very high malignancy potential of this disease.13 The colon of these patients usually appears as a carpet of polyps. These patients require colectomy by 10 to 20 years of age to prevent malignant degeneration, although an ileoproctostomy with repeat sigmoidoscopy and polypectomy every six months may be preferable in some patients. Patients with Gardner's syndrome (polyps with osteomas and soft-tissue tumors) also have adenomatous polyps that are premalignant. Juvenile polyps14 are usually solitary, but they may be present in large numbers. These can also be removed through the sigmoidoscope by an experienced sigmoidoscopist for diagnosis or for treatment of rectal bleeding. They appear inflammatory on histologic examination and are not premalignant lesions. Lymphoid nodular hyperplasia of the colon appears as submucosal nodules in large numbers, at times with the changes of colitis present in the overlying tissue.15 Patients with Peutz-Jeghers syndrome have hamartomatous polyps and mucocutaneous pigmentation. These polyps are located primarily in the small bowel, are rarely found in the colon, and are seldom malignant. There are other familial syndromes, as well as numerous patients with familial polypoid lesions who do not fit any of the described syndromes, but discussion of them is beyond the scope of this article.
Figure 4. The Quinten rectal suction biopsy apparatus can be used to obtain rectal biopsies safely from infants and small children.
Patients with chronic recurrent abdominal pain of childhood deserve special mention. Proctosigmoidoscopic abnormalities in these patients have been described,16 but all abnormalities noted, such as hyperemia or pallor, are subject to great variation in interpretation. For this reason, these patients are best diagnosed by their clinical histories and not by these findings; their proctosigmoidoscopic examinations would be considered normal according to the above-mentioned criteria.
Rectal biopsy is an essential adjunct to proctosigmoidoscopy if colitis is a diagnostic consideration. Many patients have microscopic evidence of colitis in macroscopically normal mucosa.3 Rectal biopsies can be taken either with grasp forceps through the sigmoidoscope or by a separate suction-biopsy apparatus. The advantage of doing a grasp biopsy is the ability to biopsy discrete lesions under direct observation. Unfortunately, grasp-biopsy instruments are difficult to maneuver through the smaller pediatric sigmoidoscopes, and the view is often obscured by the forceps to the extent that biopsy by this technique is hazardous with the smaller instruments. Small grasp-biopsy forceps for use with pediatric sigmoidoscopes are not yet available.
Rectal suction biopsy using the rectal biopsy attachment of the Quinton multipurpose biopsy tube17 allows safe and adequate biopsies to be taken, even in the smallest infants (Figure 4).3 The major disadvantage of this technique is that it is a blind procedure, making it impossible to biopsy discrete lesions. It is useful if one is attempting to exclude colitis or Hirschsprung's disease;1821 its diagnostic accuracy can be improved by taking several biopsies.
The use of a biopsy capsule with a 2.0-mm. infant biopsy port provides adequate-size biopsies, even in older children, and allows an additional margin of safety. To avoid risk of perforation into the peritoneal cavity, the biopsy port is best directed posteriorly or laterally because the anterior reflection of the peritoneal cavity is low. Biopsies should initially be attempted using 3-5 mm. Hg negative pressure; rarely is pressure up to 10 mm. Hg required to obtain adequate biopsies.
In addition to its role in diagnosing or excluding colitis, rectal biopsy is of use in the diagnosis of Hirschsprung's disease. The fact that suction or grasp biopsies are adequate to diagnose or exclude Hirschsprung's disease is well documented in the literature.1821 Both the superficial and deep submucous plexus and the myenteric plexus of ganglion cells are absent in all cases of Hirschsprung's disease and are always present in normal patients. Adequate samples of submucosa are usually obtained with suction -biopsy technique. If these biopsies are taken 2 cm. cephalad to the mucocutaneous junction and are serially sectioned and completely examined by a trained observer, ganglion cells can easily be found in the normal patient, excluding the diagnosis of Hirschsprung's disease. Full-thickness biopsies are therefore not necessary, eliminating the undue hazard of general anesthesia. It is important to remember, however, that the submucous plexus may be absent in the distal 1.7 cm. of the rectum, so the biopsy instrument must be inserted at least 2 cm., and more than one biopsy should be obtained.
In the preceding discussion, an introduction to the use of proctosigmoidoscopy and rectal biopsy in the pediatric patient has been presented. Greater use of these techniques in infants and children results in more precise diagnosis and treatment of disorders of the anus, rectum, and colon.
1. Baron, J. A., Cornell, A. M., and Leonard Jones, J. E. Variations between observers in describing mucosal appearances in proctocolitis. Br. Med. J. 1 (1964), 89.
2. Walts, J. M., Thompson, H., and Goligher, J. C. Sigmoidoscopy and cytology in the detection of microscopic disease of the rectal mucosa in ulcerative colitis. Gut 7 (1966), 288.
3. Vanderhoof, J. A., and Ament, M. E. Proctosigmoidoscopy and rectal biopsy in infants and children. J. Pediatr. (in press).
4. Curtis, K. J., and Sleisinger, M. H. Infections and parasitic diseases. In Sleisinger, M. H., and Fordtran, J. S. (eds). Gastrointestinal Disease: Pathophysiology, Diagnosis, Management. Philadelphia: W. B. Saunders Company, 1973, p. 1375.
5. Tuinor, J. A., et al. Amebiasis - A symposium, Calif. Med. 114 (1971), 44.
6. Helper, M-, and Ament, M. E. Shigellae dysenteriae. I: A forgotten cause of pseudomembranous colitis. J. Pediatr. (in press).
7. GroJ), A., et ai. Fulminating noninfective pseudomembranous colitis. Gastroenterology 58 (1970), 88.
8. Man, J. J., Sans, M. D.. and Tedesco. F. J. Bacterial studies of ciindamycin-associated colitis: A preliminary report. Gasfroenrerotogy 69 (1975), 342.
9. Christie, D. L., and Ament, M. E. Ampicillinassociated colitis. J. Pediatr. 87 (1975), 657.
10. Freir, S.. et al. Intolerance to milk protein, J. Pediatr. 75 (1969), 623.
11. Ament, M. E., and Rubin, C. E. Soy protein - Another cause of the flat intestinal lesion. Gasfroenterotogy 62 (1972). 227.
12. Ament, M. E. Inflammatory disease of the colon: Ulcerative colitis and Crohn's disease. J. Pediatr. 86 (1975), 322.
13. Calabro, J. J. Heritable multiple polyposis syndromes of the gastrointestinal tract. Am. J. Med. 33 (1967), 276.
14. Veale, A. M.. et al. Juvenile polyposis coli. J. Med. Genet. 3 (1966), 5.
15. Gruenberg, J., and Mackman. S. Multiple lymphoid polyps in familial polyposis. Ann. Surg. 175 (1972), 552.
16. Stone, R. T., and Barbero, G. J. Recurrent abdominal pain in childhood. Pediatrics 45 (1970), 732.
17. Brandborg. L. L,. Rubin, C. E., and Quinton, W. E. A multipurpose instrument for suction biopsy of the esophagus, stomach, small bowel, and colon. Gastroenterology 37 (1959), 1.
18. Campbell, P. E., and Noblett, H. R. Experience with recta) suction biopsy in the diagnosis of Hirschsprung's disease. J. Pediatr. Surg. 4 (1969), 410.
19. Dobbins, W. O.. and Bill, A. H., Jr. Diagnosis of Hirschsprung's disease excluded by suction rectal biopsy. N. Engl. J. Med. 272 (1965), 990.
20. Aldridge, R T., and CampbeN, P. E. Ganglion cell distribution in the normal rectum and anal canal. J. Pediatr. Surg. 3 (1968), 475.
21. Morson, B. C. Rectal biopsy in inflammatory bowel disease. N. Engl. J. Med. 287 (1972), 133.