The symptoms of inflammatory bowel disease (IBD) in children may be nonspecific and the diagnosis often delayed, especially in small intestinal Crohn's disease (1). The diagnosis of IBD conventionally depends on barium contrast radiology and colonoscopy with biopsy. In patients with small intestinal Crohn's disease, however, it is often not possible to obtain a biopsy to confirm the diagnosis histologically. Accurate diagnosis then relies on the barium contrast study. Although barium imaging and colonoscopy are important in establishing the diagnosis of IBD, they may not be the most appropriate initial investigations to use in excluding its diagnosis.
Radionuclide white cell scanning, initially with indium (2,3), and subsequently with technetium (4-6), has been use in detecting IBD in adults and may be potentially useful as a screening investigation (7). Results in two recent studies have also shown the technetium white cell scan (WCS) to be useful in the diagnosis of IBD in children (8,9).
We therefore undertook a retrospective study to assess the value of the WCS in detecting disease activity in children with active IBD, comparing it with a small bowel barium examination and colonoscopy; and to assess the use of the WCS as an initial imaging investigation in children with suspected IBD.
SUBJECTS AND METHODS
Since 1992, all children clinically suspected to have IBD have had a WCS performed as part of their diagnostic workup. This included children with such symptoms as abdominal pain, disturbed bowel habit, growth disturbance, and oral ulceration. We studied all children with known or suspected IBD who had had a WCS performed as part of their evaluation between October 1992 and November 1995 at the University Hospital of Wales, Cardiff. Subjects were identified from clinical records and the WCS records maintained in the department of radiology.
The study group comprised 55 children with proven or suspected IBD, in whom 60 WCSs were performed. Of these, 42 patients underwent WCSs initially for suspected IBD, and the other 13 were children with known IBD who were investigated at the time of clinical relapse. There were 25 male and 30 female subjects, and the median age at the time of scanning was 12.1 years (range, 1.5-18 years). Tables 1-3 show the details of patients with suspected IBD who were scanned at initial presentation and were subsequently confirmed to have IBD (7), patients with known IBD who were scanned at recurrence of symptoms (13), and patients without IBD (35).
Patients were evaluated for possible IBD on the basis of clinical suspicion or if referred for exclusion of a diagnosis of IBD. All patients with possible IBD were assessed with a WCS. The final diagnosis in all patients with IBD was made by clinical, haematologic and biochemical assessment, as well as conventional investigations, which included a barium constrast study, and total colonoscopy with histologic study on specimens from all patients. Those with known IBD (n = 13) were considered to have active disease on the basis of a combination of evidence that included recurrence of symptoms associated with elevated inflammatory indices (erythrocyte sedimentation rate, C-reactive protein, or thrombocytosis) and results of other imaging investigations (Table 2). Of the 42 who were evaluated when first seen, the 7 with a final diagnosis of IBD underwent barium contrast studies and colonoscopy as well as a WCS (Table 1). In patients without IBD (n = 35), during the initial investigation the diagnosis was made either by exclusion or by other appropriate studies (Table 3). Fourteen patients underwent a colonoscopy with biopsy and barium study, a further 10 patients had a barium study performed, and 11 patients had neither a colonoscopy nor a barium study performed. In the latter group the final diagnosis was based on the overall clinical picture and conclusive findings from other investigations.
The results of the WCS were compared with those of colonoscopy and histologic study in 7 patients with IBD with large intestine involvement, and with the results of the the barium study in 8 patients with active small intestinal disease. All investigations were performed within the same week. We then assessed the results of the WCS as an initial investigation in the 42 patients seen as new referrals in whom it was performed as part of their initial evaluation. Finally, during their assessment, the children were asked which of the investigations undertaken they disliked least.
Technetium White Cell Scanning
A 20- to 40-ml sample of venous blood was collected from the patient and the leukocytes separated and labelled under sterile conditions using technetium Tc 99m hexamethyl propylenamine oxime (HMPAO; Ceretec, Amersham International; Buckinghamshire, UK) as previously described (4,10). Anterior and posterior views of the abdomen and a pelvic outlet view were obtained 1 and 4 hours after reinjection of cells (11-13), using a large-field gamma camera, and a highresolution collimator. A delayed 24 hour image was also performed if there was doubt about the anatomic location of any abnormal activity seen on the 1- and 4-hour images. The normal colonic activity that is seen on a 24-hour image due to physiologic gut excretion allows this to be used as a template to locate abnormalities on earlier images. All scans were interpreted by an experienced nuclear medicine physician (JISR). Results were considered positive if any gut-related activity was seen on the 1-hour images. In the majority of patients the WCS was done before other investigations, and the nuclear medicine physician did not know the final diagnosis or the results of other investigations.
In the first group, which comprised patients who were subsequently confirmed to have IBD (Table 1), the WCS results were positive in all 7 patients. In 13 patients with known IBD (Table 2), 18 scans were performed during recurrence of symptoms. Of these, 17 were performed at the time of active disease; all results were positive. One result was negative in a child with an inactive stricture. This scan was performed because it was not clear whether the patient's symptoms were related to exacerbation of Crohn's disease or to obstruction secondary to stricture formation. During surgery, an inactive ileal stricture was confirmed. In the 35 patients without IBD (Table 3) who had WCSs, 3 had positive results and 32 had negative results.
Diagnosis of Active Disease
Of a total of 60 WCSs performed, 24 were done at the time of active IBD (7 in the first group, and 17 in the second group), all of which yielded positive results. Of results in patients without IBD (35 in the third group), and in the 1 patient with an inactive stricture (total 36), 33 were negative, and 3 were positive. These figures indicate that the positive WCS result has a 100% (24 of 24 in the group with active disease) sensitivity and a 91% (33 of 36 in the group with no known IBD) specificity in the diagnosis of active IBD. Overall, 3 patients had a positive WCS result (Table 3, patients 21-23) but did not have IBD. However, two of them had intestinal inflammation secondary to infectious colitis, and in only one was there no evidence of other pathology, even with laparoscopic investigation. The final diagnosis in this patient was anorexia nervosa.
Comparison of Diagnostic Methods
Among patients with large intestine disease, seven had a WCS and a colonoscopy performed within the same week. In all seven, both the WCS and histologic examination yielded positive results in the presence of similar extent of disease, giving a 100% concordance between the two.
Similarly, in the eight patients with active small intestine disease who underwent a WCS and a barium study within a few days, both methods showed disease in all eight, again with a 100% concordance between the two.
Of the 42 children evaluated with WCSs at initial examination (patients in Table 1 and 3), only 7 (17%), had IBD as a final diagnosis, and in all 7 the WCS results were positive. In the 35 patients without IBD, the WCS results were negative in 32 and positive in 3. This gave a positive WCS result a 100% sensitivity (7 of 7 patients; 95% confidence interval; range, 59-100%) (14), and a 91% specificity (32 of 35 patients; 95% confidence interval; range, 77-98%) (14), in the diagnosis of IBD.
All patients when questioned at the time of assessment, stated that they found the procedure for performing the WCS less uncomfortable and more acceptable than either the procedure for a barium study or that for a colonoscopy.
Because many of the clinical features of IBD, in particular those of Crohn's disease, are nonspecific, diagnosis of IBD must be considered in children with many different clinical symptoms, who may have a wide variety of final diagnoses, as seen in our patients in Table 3. These symptoms include recurrent or chronic abdominal pain, diarrhoea with or without blood, symptoms of anorexia, lethargy, or weight loss, growth failure, mouth ulcers, or perianal lesions. In as much as IBD is an uncommon condition in childhood, with a prevalence of 16.6 and 3.42 cases per 100,000 for Crohn's disease and ulcerative colitis, respectively, in South Wales (15), the majority of children with chronic gastrointestinal symptoms mimicking IBD, will not have IBD. This is well demonstrated in our group of 42 patients (Tables 1 and 3), in whom a final diagnosis of IBD was made in only 17% (7 of 42 patients). Clearly, in patients with possible IBD, performing a barium study and colonoscopy to confirm or to exclude the diagnosis is time consuming, expensive, and uncomfortable for the patient. The barium study, though an easy procedure to perform, imparts a significant radiation dose to a child, and the films may be difficult to interpret because of the numerous overlapping loops of small bowel. Occasionally, a small bowel enema (intubated small bowel study) is used for imaging the small bowel, but this is more invasive, in that a nasogastric tube is passed as far as the duodenojejunal flexure to give better views of the small bowel. The large bowel is ideally assessed by a colonoscopy, which is invasive, requires technical expertise, and is not widely available in children. In these circumstances a barium enema is frequently performed. This is an uncomfortable procedure and gives a higher radiation dose than a barium meal and follow-through examination. Because of the above problems of investigating children with suspected IBD with a barium study or colonoscopy, a high clinical threshold is often required for investigating children with possible IBD, especially considering its low prevalence. The high clinical threshold for investigating such children accounts for the relatively high prevalence of IBD in our series. The explanation for the high prevalence in our study may also be that many of the referrals were tertiary referrals from other consultant paediatricians who would have selected only those patients in whom IBD was more likely. However, in that the incidence of Crohn's disease in childhood is increasing (15), more children will need investigation to exclude the diagnosis of IBD. It is thus important to have a relatively less invasive test to exclude IBD.
Results of numerous studies in adults, reviewed recently by Weldon (13) and Giaffer (16), have confirmed the benefit of the WCS in the diagnosis of IBD. White cell scanning has also been more sensitive than a barium study in detecting IBD (2). Bhargava et al. (8) and Jobling et al. (9) have reported on the usefulness of the WCS in children with IBD. Our data suggest that the WCS is the best initial imaging investigation (if facilities for labelling are available) in children suspected to have IBD. As a screening test, it has many advantages over the barium study and the colonoscopy. Its potential benefits are that it is less uncomfortable and more acceptable to the patient, that it can assess both the small and large intestines simultaneously, that it is noninvasive and can easily be performed on a sick patient, and that it subjects the patient to a significantly lower radiation exposure than a barium study. The cost of a WCS is comparable to that of a barium study but less than that of a colonoscopy; and as an initial investigation it is considerably cheaper than a barium study and colonoscopy performed together, in as much as both these latter tests should be done to diagnose or exclude IBD, particularly Crohn's disease.
Recently, Beattie et al. (17) have suggested that acute inflammatory indices be used to select patients for further investigation. However, in our series four patients (patients 22,45,47, and 49 in Table 3) had elevated inflammatory indices without IBD. Only one of these had a positive WCS result and thus would have undergone further evaluation, instead of all four as suggested by Beattie et al. Additionally, on five occasions in patients with known relapse of IBD, the WCS results were positive but acute inflammatory indices were normal. The diagnosis of relapse in these patients was based on the overall assessment of the patient, which included recurrence of symptoms at least 1 year after initial diagnosis with barium study or WCS results showing a different site of involvement. This suggests that a WCS may be a better discriminator than acute inflammatory indices.
The WCS is easily performed in children as young as 1.5 to 2 years, from whom approximately 20 ml of blood must be aspirated for white cell separation and labelling, before reinjection. Imaging time is approximately 15 minutes for each visit (at 1 and 4 hours after injection) and does not cause the patient any discomfort. Sedation is rarely required. The labelling procedure takes approximately 1 to 1.5 hours and has to be done under strict aseptic conditions; if appropriate facilities are not available, it may require a small capital outlay by a hospital newly undertaking the technique.
In our experience in the use of the WCS as an initial screening investigation, there were no false-negative results. Therefore in the group of 42 children with suspected IBD who underwent a WCS at initial examination (patients in Tables 1 and 3), 32 patients with negative results (76%) would not have required further evaluation with a barium study or a colonoscopy. It is now our policy not to perform these investigations in children with negative WCS results, unless there remains a strong clinical suspicion of IBD.
As mentioned earlier, of the three patients with a positive WCS result without IBD, two had other inflammatory conditions of the gastrointestinal tract. As experience with WCS increases, there is a growing list of false-positive results (18-24) that includes uptake in stoma sites and wounds, tumours (carcinoma and lymphoma), haematomas and gastrointestinal haemorrhage, deep vein thrombosis, vasculitis, ischaemic and infective bowel disease, pseudomembranous colitis, intramuscular injections, and swallowed activity from chest infections. The commonest problem is the distinction from the normal physiologic gut activity usually seen in the right iliac fossa from 2 to 3 hours after reinjection. This is why an image taken at 1 hour after injection is essential (13). Accurate scan interpretation relies on a full recognition of all the above possibilities, particularly in a group of children undergoing initial investigations. A positive WCS result should be seen as an initial investigation in this group, requiring further confirmation by radiologic and histologic study.
Our results are comparable to those in recently published studies in children (8,9) and show that the WCS has a very good correlation with active IBD in children, and compares favourably with a barium study and a colonoscopy. The WCS can be used for the rapid assessment of acutely ill patients with suspected fulminant colitis (13), in whom a colonoscopy may be too hazardous, and a plain abdominal film provides inconclusive results. The WCS, although not essential, may also be useful in some difficult cases of relapse and may complement other investigations-for example, those assessing the activity of strictures identified by barium studies-or it may show a different unsuspected site of involvement that may require modification of treatment.
In our experience, by following the strategy of investigating all patients with suspected IBD with a WCS initially, and only subjecting those with a positive result to a colonoscopy to obtain histologic confirmation of the diagnosis, no diagnosis of IBD would be missed, and the majority of children would not have to undergo an invasive procedure. Obviously, there must remain a place for clinical suspicion, in that it is unlikely that any test will be 100% sensitive when used more extensively, but the number of cases of active IBD with negative WCS results should be very small.
In summary, our results confirm the value of WCS in assessing disease activity and extent in the follow-up of patients with known IBD. They also suggest that the WCS may be very useful as an initial imaging investigation in the assessment of children with suspected IBD, to select patients who need further evaluation with contrast radiology and endoscopy to obtain tissue for histologic confirmation of the diagnosis. The latter may include upper gastrointestinal endoscopy or colonoscopy, or both, depending on the distribution of abnormal uptake shown on the WCS.
Acknowledgment: The authors thank Mr. J. Jones for his radiopharmacy support and advice.
1. Burbige EJ, Huang SS, Bayless TM. Clinical manifestations of Crohn's disease in children and adolescents. Pediatrics
2. Saverymuttu SH, Peters AM, Crofton ME, et al. 111
Indium granulocytes in the detection of inflammatory bowel disease. Gut
3. Saverymuttu SH, Camilleri M, Rees H, Lavender JP, Hodgson HJF, Chadwick VS. Indium 111-granulocyte scanning in the assessment of disease extent and disease activity in inflammatory bowel disease. A comparison with colonoscopy, histology, faecal indium 111-granulocyte excretion. Gastroenterology
4. Peters AM, Danpure HJ, Osman S, et al. Clinical experience with 99m
Tc-hexamethylpropylene-amine oxime for labelling leucocytes and imaging inflammation. Lancet
5. Becker W, Schomann E, Fischbach W, Borner W, Gruner KR. Comparison of 99
-HMPAO and 111
In-Oxine labelled granulocytes in man: first clinical results. Nucl Med Commun
6. Roddie ME, Peters AM, Danpure HJ, et al. Inflammation: Imaging with Tc-99m HMPAO-labeled leukocytes. Radiology
7. Gibson P, Lichtenstein M, Salehi N, Hebbard G, Andrews J. Value of positive technetium-99m leucocyte scans in predicting intestinal inflammation. Gut
8. Bhargava SA, Orenstein SR, Charron M. Technetium-99m hexamethyl-propylenamine-oxime-labelled leukocyte scintigraphy in inflammatory bowel disease in children. J Pediatr
9. Jobling JC, Lindley KJ, Yousef Y, Gordon I, Milla PJ. Investigating inflammatory bowel disease-white cell scanning, radiology, and colonoscopy. Arch Dis Child
10. Kelbaek H, Linde J, Nielson SL. Evaluation of a new leukocyte labelling procedure with 99m
Tc-HMPAO. Eur J Nucl Med
11. Kennan N, Hayward M. Tc-HMPAO-labelled white cell scintigraphy in Crohn's disease of the small bowel. Clin Radiol
12. Weldon MJ, Grundy A, Joseph AEA, Maxwell JD. Comparison of Technetium (Tc99m) HMPAO white cell scanning with barium radiology in the assessment of small bowel Crohn's disease. Gut
13. Weldon MJ. 99m
Tc-HMPAO planar white cell scanning. Scand J Gastroenterol
14. Lentner C, ed. Geigy Scientific Tables
. 8th ed. Basel, Switzerland: Geigy, 1982:89-102.
15. Cosgrove M, Al-Atia RF, Jenkins HR. The epidemiology of paediatric inflammatory bowel disease. Arch Dis Child
16. Giaffer MH. Labelled leucocyte scintigraphy in inflammatory bowel disease: Clinical applications. Gut
17. Beattie RM, Walker-Smith JA, Murch SH. Indications for investigation of chronic gastrointestinal symptoms. Arch Dis Child
18. Lantto EH, Lantto TJ, Vome M. Fast diagnosis of abdominal infections and inflammations with technetium-99m-HMPAO labeled leukocytes. J Nucl Med
19. Peters AM. Imaging inflammation: Current role of labeled autologous leucocytes (editorial). J Nucl Med
20. Lantto E. Investigation of suspected intra-abdominal sepsis: The contribution of nuclear medicine. Scand J Gastroenterol
21. Figa M, Acero D, Gonzalez-Huix F, Berna L. Tc-99m HMPAO labeled leucocyte uptake by an inflammatory fibroid polyp of the ileum. Clin Nucl Med
22. Evetts BK, Foley CR, Latimer RG, Rimkus DS. Tc-99 hexamethylpropylenamineoxide scanning for the detection of acute appendicitis. J Am Coll Surg
23. Nathan MA, Seabold JE, Brown BP, Bushnell DL. Colonic localization of labeled leucocytes in critically ill patients. Scintigraphic detection of pseudomembranous colitis. Clin Nucl Med
24. Al-Nahhas AM, Hjiyiannakis P, Hammersley PA, McCready VR, Horwich AH. Uptake of technetium-99m hexamethylpropylene amine oxime labelled white cells in lymph nodes in non-Hodgkin's lymphoma. Eur J Nucl Med