Thomson, Mike; Kitching, Paul; Jones, Alison; Walker-Smith, John A.; Phillips, Alan
Background: Endoscopy is increasingly used in paediatric practice for diagnosis of enteropathy, although the quality of grasp forceps-obtained biopsy specimens required for reliable diagnosis has been questioned in comparison with suction capsule biopsy specimens. This study prospectively compared the diagnostic suitability of grasp forceps biopsy versus suction biopsy in the same patient during the same procedure.
Methods: A double-port paediatric suction biopsy capsule was front-loaded onto an endoscope and directed to the fourth part duodenum-proximal jejunum for biopsy sampling. Subsequently, three grasp biopsy specimens were taken from the same region. All biopsies were coded, photographed, and measured for area, using computed morphometry. A single blinded histopathologist assessed sample adequacy for diagnosis. Twenty-nine patients were enrolled (age range, 8-185 months).
Results: On three occasions the suction capsule failed to fire, and on four occasions only one sample was obtained. Three grasp biopsy specimens were obtained on each occasion by endoscopy, and the first two were used for comparison with suction biopsy samples. Median total area of individual biopsy samples obtained by the two procedures was not different (21.3 vs. 22.5 mm2; P = 0.027). Muscularis mucosae was obtained more commonly with grasp biopsies (P < 0.001), and no difference was observed for the presence of three or more villuscrypt units, degree of haemorrhage, or optimal orientation. Two suction biopsy procedures and one grasp biopsy procedure were inadequate for diagnosis.
Conclusions: Endoscopic grasp biopsies are perfectly adequate for the assessment of small intestinal histology. In addition, endoscopy affords advantage in diagnosis of other upper gastrointestinal disease with avoidance of radiologic screening involved with the suction capsule technique.
The identification of small bowel disease in infants and children with suspected small bowel enteropathy continues to rest on biopsy and histologic analysis of the small bowel. Classically, the diagnosis of coeliac disease rests on the evidence obtained in small intestinal biopsy, best described as hyperplastic villous atrophy with hyperplasia of the crypts and an abnormal surface epithelium (1). Other diseases such as cow's milk-sensitive enteropathy (CMSE) and autoimmune enteropathy may involve patchy lesions that could be missed by use of a small bowel biopsy capsule with a single port (2,3). The double-port capsule was developed to take account of this patchiness (4). However, endoscopic biopsy has a great advantage in its capacity to obtain multiple mucosal biopsy specimens. Furthermore, the capsule biopsy method has the disadvantage of requiring radiologic screening in young patients (5). Sullivan et al.(6) describe a method for front-loading a small bowel biopsy capsule onto an endoscope and guiding the capsule into the duodenum avoiding irradiation-involving endoscopy. Historically, the capsule technique has had a failure rate of between 10% and 20%, and provided specimens that were inadequate for making a histologic diagnosis in up to 25% of cases (7,8) and in more recent articles, between 6% and 19% (9,10).
Endoscopy has become more widely used with the need to study macroscopic and microscopic disease of the stomach and oesophagus in addition to the small intestine. The endoscopic technique of sampling tissue is limited by the size of the biopsy channel and therefore of the grasp biopsy forceps, and concern has been raised about the quality of the samples obtained-albeit offset by the ability to perform multiple biopsies during the procedure.
Comparison of the biopsy's adequacy for diagnosis has been made by several groups (9-13); however, these have been mainly retrospective and have comprised series performed in different patients at different times, and the size of the forceps has not been uniform. It is possible to front-load a suction capsule on an endoscope (6), and in this study, we adopted this technique to conduct a prospective study comparing suction biopsy and grasp biopsy samples obtained from the same patient, during the same procedure, by a single endoscopist.
PATIENTS AND METHODS
Twenty-nine children (median age, 35 months; range 8-185 months) coming consecutively to the Department of Gastroenterology for evaluation of suspected malabsorption and small bowel enteropathy (including suspected coeliac disease, CMSE, and autoimmune enteropathy) underwent small bowel biopsy by endoscopic grasp biopsy forceps technique (E) and small bowel biopsy capsule (C). Fully informed parental consent was obtained, and the endoscopy was performed by the same operator under a short general anaesthetic, as is normal practice for upper endoscopy in children under 5 years of age in this unit. A double-port Crosby small bowel biopsy capsule was front-loaded onto the video endoscope (XQ230; Olympus, Tokyo, Japan) through the biopsy channel (2.8 mm), according to the technique of Sullivan et al., (6) and manoeuvred into the distal duodenum or proximal jejunum under direct vision. All insufflated air was extracted, the capsule was fired, and the endoscope and capsule were removed. If the capsule had not successfully fired, this was noted as a failure, and it was not reinserted. Subsequently, three specimens were obtained by endoscope using grasp biopsy forceps (FB25K-1; Olympus) under direct vision from the same portion of duodenum or jejunum from which the capsule specimens had been obtained. All specimens were randomly numbered, oriented immediately in the endoscopy suite under a dissecting microscope (Motic SMZ-143 stereo zoom microscope with OM-2 camera), photographed at a fixed magnification, and placed in formaldehyde for histology.
The surface area of all specimens was then measured by a single blinded observer using computed morphometry from the photographs taken at the dissecting microscope stage. This involved tracing the outline of the sample, and the area within this outline was calculated by the computer using commercially available software (Kompira Ltd., Scalsburgh, Scotland). Each specimen was serially sectioned and sections 11, 21, and 31 were examined; the intervening material was kept for further staining if necessary. All were examined with haematoxylin and eosin staining and some, but not all, with diastase periodic acid-Schiff staining. Histologic examination by a single blinded histopathologist for the criteria noted in Table 1 was then undertaken. Orientation was considered optimal if the section was precisely perpendicular to the muscosal sheet, but not to so severe a degree that a reasonable judgment could not be made about crypt and villous sizes, taking all levels into account. Orientation was considered poor if sectioning was at so oblique an angle that crypt and villous morphology could not be assessed for diagnosis. Assessment for suitability for diagnosis took into account the overall size of the specimen, its orientation, and the presence of obscuring factors such as trauma and haemorrhage. Suitability for diagnosis was considered optimal if a diagnosis could be assigned with ease from the material; suboptimal if assessment was partially compromised allowing only a partial assessment to be made, perhaps preventing exclusion of subtle or patchy lesions; and inadequate if the material was too scanty, too poorly oriented, or too traumatized for assessment.
Student's paired t-test, analysis of variance, and Wilcoxon signed rank test were used, statistical significance was taken at P < 0.05. The study received approval from the Royal Free Hospital Ethics Committee.
The diagnoses obtained were as follows: six coeliac disease, five CMSE, and three autoimmune enteropathy; eight had normal small bowel histology.
In three cases the suction biopsy capsule did not fire, and in four, only 1 specimen was obtained from the dual-port capsule. Therefore, comparison of specimens was not possible in 7 cases, leaving 22 in which comparison of the two techniques was attempted.
Figure 1 shows a multiple box-and-whisker plot for the average size of the individual biopsy samples, and it can be noted that neither the mean nor median surface area was statistically different, although the specimens obtained by capsule display a wider range. This still holds if the first two endoscopic specimens rather than all three are compared with the two specimens obtained by capsule.
Table 2 summarizes the histologic information for comparable groups: two capsule specimens versus the first two endoscopic specimens obtained. It can be seen that, apart from trauma artefact and depth of biopsy, there was no statistically significant difference in other parameters. The endoscopic specimens included muscularis mucosae in 10 of 22 cases, which was not seen in any capsule specimens. Trauma artefact was more common in the endoscopic biopsy group (P = 0.01), but this did not affect the potential for making a diagnosis. The specimens were deemed inadequate for reliable diagnosis in 2 of 22 and 3 of 22 capsule and endoscopic cases, respectively; however, in 7 additional cases, the capsule did not fire or only obtained one specimen, which precluded size and histologic comparison. In no case was there a failure to obtain at least three endoscopic specimens.
This study shows that small bowel mucosal samples obtained by the endoscopic technique with grasp biopsy forceps are as adequate as those obtained by the suction capsule technique for the exclusion or confirmation of small bowel enteropathy in paediatric practice, are of a similar (although more consistent) size to suction-obtained specimens, and have a 100% success rate compared with 76% with the capsule. This guarantee of obtaining specimens is particularly important in paediatrics in which it is optimal if repeated exposure to irradiation (5) or to sedation-general anaesthetic can be minimised. This study is unique in paediatrics, in that we attempted to obtain specimens from the same patients, and furthermore from the same intestinal site with both techniques, which has only been achieved previously in one study in adults (11). No complications were noted with either technique, although the airway was protected and no radiograph screening was used, neither of which are normally the case when biopsies are performed with the capsule technique.
The artefact rate was higher in the endoscopic biopsies but this did not adversely affect the ability to make or refute a histologic diagnosis. Similarly, the proportion of biopsies containing muscularis mucosae was higher in the endoscopic group with no effect noted on diagnostic precision. The importance of immediate and accurate orientation of the specimens with the aid of a dissecting microscope is strongly emphasised, as has been pointed out previously (8,14-16). The advantages of the endoscopic technique are added to when it is considered that numerous specimens can be obtained from the small bowel by this method (this study dealt with the arbitrary but practical comparison of two capsule specimens, when this was successful, with two or three endoscopic specimens).
It must be emphasised that the standard (Olympus FB25K-1), rather than the paediatric (FB19K-1), biopsy forceps should be used when possible-these are insertable into an endoscope that can be used in patients as young as 6 months (Olympus XQ230). It is notable that those who have claimed that endoscopic biopsies are less reliable have not used standard-sized biopsy forceps (8).
In addition, endoscopy allows greater flexibility for biopsy from different sites from the small bowel, having greater potential to show the patchy nature of the enteropathy in some conditions in infancy and childhood, such as CMSE. Postmortem studies (15) and double-port capsule studies have shown that noncoeliac enteropathies are typically patchy (17). In addition, endoscopy allows visualisation of the oesophagus, stomach, and duodenum for the presence of macroscopic disease (e.g., oesophagitis associated with CMSE, or upper gastrointestinal Crohn's disease) (18). Previous studies have been concerned with the small size of the endoscopic specimens, but surface area of the specimens have not been objectively measured (3,11,12,19). There was no statistically significant difference in the sample size, when measured by computed morphometry of photographs of the specimens in our study. It was noted by other groups that if there were inadequacy of sample size as assessed by dissecting microscope then further endoscopic biopsies could be performed at the same time (12). Indeed, we have found that the combination of dissecting microscopy with endoscopic biopsy affords flexibility of response to maximise efficiency and greatly improves the accuracy of diagnosing enteropathy (14,15). Hypoxia is a frequent occurrence in upper oesophagogastroduodenoscopy under sedation, and general anaesthesia is a safer technique that prevents hypoxia and allows the operator to perform the endoscopy under better conditions (20).
In 1990, a working group of the European Society of Paediatric Gastroenterology, Hepatology, and Nutrition suggested that, in the case of coeliac disease, a small intestinal biopsy should always be performed with a suction biopsy capsule to ensure diagnostically adequate specimens (1). However, the results reported in the current study, involving the same consecutive subjects with specimens taken from the same intestinal area by both techniques, would strongly suggest that duodenal-jejunal biopsy through the endoscopic technique using adultsized 2.4-mm grasp forceps is an acceptable alternative, with the extra advantage of allowing as many biopsies to be performed as are needed. This is also important for research purposes in diseases of uncertain pathogenesis.
In conclusion, this study shows that the better method for sampling of the small bowel mucosae in infants and children is by the endoscopic route.
1. Walker-Smith J, Guandalini S, Schmitz J, Schmerling D, Visakorpi J. Revised criteria for diagnosis of coeliac disease. Arch Dis Child 1990;65:909-11.
2. Shiner M. Duodenal biopsy. Lancet 1956;1:17-19.
3. Collins A, Brookfield D, Hyde I, Rolles C. Small bowel biopsy. Arch Dis Child 1985;70:1082-5.
4. Kilby A. Paediatric small intestinal biopsy capsule with two ports. Gut 1976;17:158-9.
5. Granditsch G, Deutsch J, Tsarmaklis G, Kletter K. Exposure to X-rays during small bowel biopsies in children. Eur J Pediatr 1981;137:165-9.
6. Sullivan P, Phillips M, Neale G. Endoscopic capsule biopsy under endoscopic guidance. J Pediatr Gastroenterol Nutr 1988;7:544-7.
7. Smith R, Sprinz H, Crosby W, Sullivan B. Peroral small bowel mucosal biopsy. Am J Med 1958;25:391-4.
8. Scott B, Jenkins D. Endoscopic small intestinal biopsy. Gastrointest Endosc 1981;27:162-7.
9. Eltumi M, Ong P, Francis N, Brueton M. A comparison of endoscopic and small intestinal biopsy techniques in children with upper gastrointestinal disorders. J Paediatr Child Health 1996;32:255-6.
10. Branski D, Faber J, Freier S, Gottschalk-Sabag S, Shiner M. Histological evaluation of endoscopic versus suction biopsies of small intestinal mucosae in children with and without celiac disease. J Pediatr Gastroenterol Nutr 1998;27:6-11.
11. Mee A, Burke M, Vallon A, Newman J, Cotton P. Small bowel biopsy for malabsorption: Comparison of the diagnostic adequacy of endoscopic forceps and capsule biopsy specimens. BMJ 1985;291:769-72.
12. Gottrand F, Turck D, Mitchell V, Farriaux J. Comparison of fibre endoscopy and Watson capsule for small intestinal biopsy in infants and children. Acta Paediatr 1992;81:399-401.
13. Granot E, Goodman-Weill M, Pizov G, Sherman Y. Histological comparison of suction capsule and endoscopic small intestinal mucosal biopsies in children. J Pediatr Gastroenterol Nutr 1993;60:397-401.
14. Walker-Smith J. Dissecting microscope appearance of small bowel mucosae in children. Arch Dis Child 1967;42:626-30.
15. Walker-Smith J. Uniformity of dissecting microscope appearances in proximal small intestine. Gut 1972;13:17-20.
16. Gillberg R, Kastrup W, Mobacken H, Stockbrugger R, Ahren C. Endoscopic duodenal biopsy compared with biopsy from the upper jejunum in patients with dermatitis herpetiformis. Scand J Gastroenterol 1982;17:305-8.
17. Manuel P, Walker-Smith J, France N. Patchy enteropathy. Gut 1979;20:211-5.
18. Barkin J, Schonfield W, Thomsen S, Manten H, Rogers A. Enteroscopy and small bowel biopsy-an improved technique for the diagnosis of small bowel disease. Gastrointest Endosc 1985;31:215-7.
19. Kirberg A, Lattore J, Hartard M. Endoscopic small intestinal biopsy in infants and children: Its usefulness in the diagnosis of celiac disease and other enteropathies. J Pediatr Gastroenterol Nutr 1989;9:178-81.
20. Lamireau T, Dubreuil M, Daconceicao M. Oxygen saturation during esophagogastroduodenoscopy in children: General anaesthesia versus intravenous sedation. J Pediatr Gastroenterol Nutr 1998;27:172-5.
© 1999 Lippincott Williams & Wilkins, Inc.