Are 99mTc Leukocyte Scintigraphy andSBFT Studies Useful in Children Suspectedof Having Inflammatory Bowel Disease?Martin Charron, M.D., Carlo Di Lorenzo, M.D., and Samuel Kocoshis, M.D.Departments of Radiology and Gastroenterology, Children’s Hospital of Pittsburgh, andUniversity of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania

OBJECTIVE: The goal of this retrospective study was to as-sess whether99mTc-white blood cell (WBC) scintigraphyand upper gastrointestinal small bowel follow-through(UGI-SBFT) could exclude inflammation in children sus-pected of having inflammatory bowel disease (IBD).

METHODS: Of a population of 313 children who had a99mTc-WBC scan, 130 children were studied exclusively torule out IBD. Sixty-nine colonoscopies with biopsies weredone within a short time interval of the99mTc-WBC scans.There were also 51 controls studied with99mTc-WBC scin-tigraphy.

RESULTS: Of the 130 children studied to exclude IBD, thefinal diagnosis was Crohn’s disease in 27, ulcerative colitisin nine, miscellaneous colitis in 13, probably normal in 42,and normal in 39. The99mTc-WBC scans were positive inall but three newly diagnosed Crohn’s disease, ulcerativecolitis, or miscellaneous colitis children. The false-negative99mTc-WBC studies were seen in children with mild inflam-mation on biopsies and normal UGI-SBFT studies. In the 46children with a true-positive99mTc-WBC scan, 81% (17/21)of UGI-SBFT studies were normal. In five children withequivocal UGI–SBFT studies, the99mTc-WBC scan cor-rectly predicted if inflammation was present in the terminalileum.

CONCLUSIONS: Our results suggest that99mTc-WBC is use-ful as an initial screening modality to exclude IBD, and ismore sensitive than UGI-SBFT studies. (Am J Gastroenterol2000;95:1208–1212. © 2000 by Am. Coll. of Gastroenter-ology)

INTRODUCTION

The diagnosis of inflammatory bowel disease (IBD) in chil-dren is often complex and difficult, and the diagnosis isoften delayed by many months. For many years, radiolog-ical examinations were the cornerstone of diagnostic testing,and more recently, colonoscopy has been used more liber-ally. We have evaluated99mTc-white blood cell (WBC)scintigraphy in children with known IBD (1). However, ithas been suggested that in patients with symptoms sugges-

tive of IBD that 99mTc-WBC scintigraphy is poor as ascreening test (2). Therefore, the purpose of this study wasto determine, compared with colonoscopy and biopsies,whether99mTc-WBC scintigraphy and small bowel follow-through (SBFT) can exclude IBD among a large populationof children screened for IBD.

MATERIALS AND METHODS

PatientsOver a 6-yr period (10/1992–12/1998), 31399mTc-WBCstudies were performed and the hospital charts of theseconsecutive children were reviewed. There were threegroups of children evaluated. The first group was made upof 130 children who had a99mTc-WBC scan to exclude IBD(RO); the second group was 132 children with known (1)IBD who had a99mTc-WBC for assessment of a clinicalflare. The accuracy of99mTc-WBC scintigraphy has beenevaluated in this latter group of children with known IBD(1), but not in children suspected of having IBD. The currentstudy includes new data from 98 children not previouslyreported (1), and data from SBFT studies not reported ineither group. There were 144 boys and 169 girls (averageage, 13 yr).

The 130 children with suspected IBD had symptomssuggestive of IBD and/or persistently abnormal laboratoryvalues. The average clinical follow-up was 350 days. Therewere 54 boys and 76 girls (average age, 12 yr). In this groupof children with suspected IBD, 42 upper gastrointestinal(UGI)-SBFT studies were available for review. Twenty-oneof these UGI-SBFT studies were performed within 3 days ofthe 99mTc-WBC scan and 21 within 2 wk.

The third group consisted of 51 controls (NL) who un-derwent99mTc-WBC scanning for other medical problemsin which the final diagnoses were osteomyelitis, cellulitis,trauma, myocarditis and fever of unknown origin, painfulhip, and peritonitis.

ColonoscopyAmong the 130 children within the RO group, colonoscopywas performed in 69 children. The average elapsed time

THE AMERICAN JOURNAL OF GASTROENTEROLOGY Vol. 95, No. 5, 2000© 2000 by Am. Coll. of Gastroenterology ISSN 0002-9270/00/$20.00Published by Elsevier Science Inc. PII S0002-9270(00)00803-0

between colonoscopy and99mTc-WBC scintigraphy was 11days (median, 5 days).

Total colonoscopy using an Olympus CF100TL or PCF20instrument (Olympus Optical, Tokyo, Japan) was performedby an experienced endoscopist to assess extent and activityof the mucosal inflammatory changes in eight intestinalsegments (defined later). The endoscopic findings of inflam-matory activity were classified as absent (noninflamed mu-cosa), mild (granularity, edema, invisible vascular pattern),moderate (hyperemia, friability, and all features of mildactivity), or severe (ulceration, in addition to features ofmoderate activity).

Performance of the ScansLabeling of the leukocytes with99mTc has been described(3). At 0.5–1 h after injection, imaging was done with agamma camera (Siemens Orbiter, Des Plaines, IL). Two tofour hours postinjection, an 8-min anterior view of theabdomen was repeated. SPECT images were acquired at2–4 h (4).

Analysis of the ScansIn each set of scans (0.5–1 h and 2–3 h) from the 313patients, the bowel was divided into eight segments (rectum,sigmoid, descending, transverse, ascending, cecum, termi-nal ileum, and small bowel), resulting in 5008 bowel seg-ments for scoring (i.e., 8 segments3 2 images3 313patients5 5008). The inflammatory activity in each seg-ment was graded semiquantitatively by comparing the up-take in the bowel with that in the iliac crest bone marrowand liver: Grade 05 no activity, Grade 15 activity lessthan iliac crest, Grade 25 activity similar to iliac crest,Grade 35 activity greater than iliac crest, Grade 45activity equal to liver, Grade 55 activity greater than liver,Grade 65 activity equal to spleen. The grade of uptake onthe99mTc-WBC scans was then compared with the colonos-copy result. All images were interpreted by one nuclearphysician who was blinded to the clinical details. The scin-tigraphic result was considered a false negative if there wasa difference in the degree of inflammation between the99mTc-WBC findings and the colonoscopic result,i.e., a99mTc-WBC scan with mild inflammation (grade 1 or 2) andsevere inflammation at colonoscopy was considered afalse negative. The scintigraphic result was interpreted asa false positive if it revealed uptake of99mTc-WBC in anarea that was normal on colonoscopy and biopsy. Activityseen at a colostomy or ileostomy site was excluded fromanalysis.

Statistical MethodsThe histological findings in each segment were comparedwith the results of the99mTc-WBC scan and the endoscopicassessment of the corresponding segment. Sensitivity, spec-ificity, positive predictive value, negative predictive value,and diagnostic accuracy were calculated, choosing the histo-logical assessment as the reference method when specified.

RESULTS

From the group of 98 children not previously reported (1),25 colonoscopic results were available. When comparedwith colonoscopy and biopsies, there were 13 true-positive99mTc-WBC studies, 11 true-negative studies, one false-negative99mTc-WBC scan, and no false-positive99mTc-WBC studies. In this small subgroup of children studiedexclusively to rule out IBD, the sensitivity of99mTc-WBCscintigraphy was 93%, the specificity was 100%, the posi-tive predictive value was 100%, and negative predictivevalue was 92%. The colonoscopy in the child with a falselynormal 99mTc-WBC scan showed a friable terminal ileum,and the biopsy specimen revealed lymphoid follicles withchronic inflammation in the ileum and prominent lymphoidaggregate with mild architectural distortion in the cecum(Fig. 1).

From the entire dataset of 130 children studied to excludeIBD, the final diagnosis (by colonoscopy and biopsies) wasCrohn’s disease (CD) in 27, ulcerative colitis (UC) in nine,miscellaneous colitis (MC) in 13 (seven indeterminate co-litis, four infectious colitis, and two autoimmune colitis),probably normal in 42 (children with no evidence of IBD byclinical follow-up and radiographic study), and normal in 39

Figure 1. Abnormal uptake of99mTc-WBC in the terminal ileum(arrow) in a child with Crohn’s disease. The uptake is intense,bowel-shaped, and appears early (30 min). On the late scan, theuptake did not move or change shape, suggesting active IBD(documented by biopsies).

1209AJG – May, 2000 99mTc Leukocyte Scintigraphy and SBFT in Children

(normal colonoscopy). Table 1 details the calculated accu-racy of 99mTc-WBC scintigraphyversuscolonoscopy. The99mTc-WBC scan suggested the presence of IBD in 36%(49/129) of children studied to exclude IBD. The99mTc-WBC scans were positive in all but two newly diagnosedCD or UC patients. These two false-negative99mTc-WBCstudies were seen in a child whose colonoscopy showed amucosa characterized by a few aphtoid ulcers, and in asecond child with UC with an edematous mucosa. Thebiopsy results showed mild chronic inflammation. The otherfalse-negative99mTc-WBC study was seen in a child with afinal diagnosis of autoimmune colitis rather than IBD. In

these three children with a false-negative99mTc-WBC scan,laboratory values and UGI-SBFT studies were normal,when available (2/3). The four false-positive99mTc-WBCstudies were seen in a child with a GI bleed, in one with adiversion colitis, and in two children with uptake in theproximal small bowel that remained unexplained and notcorroborated by other diagnostic modalities and long-termclinical follow-up.

Small Bowel Follow-ThroughIn children with true positive99mTc-WBC scans, 81% (17/21) of available UGI-SBFT studies were normal (includingtwo that were equivocal). These consisted of six normal andtwo positive UGI-SBFT done within 3 days of the99mTc-WBC scan, and 11 negative and two positive UGI-SBFTdone within 2 wk of the99mTc-WBC scan. In the 79 childrenwith a true-negative99mTc-WBC study, there were 18 true-negative UGI-SBFT studies and three equivocal studies(lymphoid nodular hyperplasiavs CD).

Of the 313 children studied with99mTc-WBC scintigra-phy, there were seven false-negative and eight false-positive99mTc-WBC studies. In the seven children with false-nega-tive 99mTc-WBC scans all available UGI-SBFT were nor-mal.

DISCUSSION

The approach to the child with suspected IBD is complexand the diagnosis is often delayed (5). Children may nothave experienced serious health problems before the onsetof IBD. Therefore, recognition of the emotional impact ofintrusive routine diagnostic studies, such as rectal examina-tions, radiographs, and endoscopy, is essential (6). Theaccuracy of colonoscopic diagnosis of IBD is in the range of80–90% (7–9). In approximately 30–40% of patients, CDcan involve the terminal ileum alone, and spare the largebowel. Stricture, spasm, and other technical factors precludethe performance of total colonoscopy and ileoscopy in ap-proximately 20–40% of patients with suspected CD (10,11). Therefore, several studies are needed to analyze theentire bowel.

There are data to document the accuracy of99mTc-WBCin children with known IBD (1), but not in children withsuspected IBD. We document in this report the accuracy of99mTc-WBC scintigraphy for the evaluation of children withsuspected IBD. We found that the99mTc-WBC study wasunlikely to miss significant inflammation when screeningchildren for IBD. In the 130 children studied to excludeIBD, there were only two false-negative99mTc-WBC stud-ies, with a third false-negative proven autoimmune colitis.These children had mild inflammation as evidenced bynormal laboratory values, mild chronic inflammation onbiopsies, and normal UGI-SBFT studies. We had no false-positive studies in the controls and this confirms the highnegative predictive value of99mTc-WBC scintigraphy. The99mTc-WBC scan was more sensitive than UGI-SBFT to

Figure 2. Planar images disclose continuous colonic uptake in onechild with ulcerative colitis. 3-D images are available for review atthe web page: HTTP://www.arad.upmc.edu/users/charron/index.htm.

Table 1. Calculated Accuracy* of 99mTc-WBC ScintigraphyVersusColonoscopy in Children With Suspected IBD

Suspected IBD

Number of colonoscopy 69False negative 3False positive 1True positive 49True negative† 67Sensitivity 94%Specificity 99%Positive predictive value 98%Negative predictive value 96%Accuracy 97%

* Sensitivity (%)5 (true positive * 100)/(true positive1 false negative); specificity(%) 5 (true negative * 100)/(true negative1 false positive); positive predictive value(%) 5 (true positive * 100)/(true positive1 false positive); negative predictive value(%) 5 (true negative * 100)/(true negative1 false positive); accuracy5 ((truepositive1 true negative) * 100)/(true positive1 false negative1 false positive1 truenegative).

† includes 46 controls.IBD 5 inflammatory bowel disease.

1210 Charron et al. AJG – Vol. 95, No. 5, 2000

screen children for IBD. In the 46 children with true-positive 99mTc-WBC scans, the UGI-SBFT studies werefalsely normal in 81% of children. In five children withequivocal UGI SBFT,i.e., lymphoid nodular hyperplasiaversus CD, the 99mTc-WBC scan correctly predictedwhether inflammation was present in the terminal ileum.

Endoscopic and radiological methods of disease localiza-tion are more invasive when compared with the99mTc-WBCscan, and tend to produce more discomfort related to instru-mentation and preparation for the procedure (e.g., bowelcleansing). Scintigraphy with99mTc-WBC has been re-ported to be sensitive for the detection of inflammation inadults (2, 12–20) and in series with small number of chil-dren (21–23). The99mTc-WBC scan seems ideally suited toobtain a precise temporal snapshot of the distribution (24)and intensity of inflammation in the large and small bowel,whereas radiographic modalities of investigation tend torepresent changes that are more chronic (25–27). Occasion-ally the 99mTc-WBC scan can help characterize equivocalfindings on the UGI-SBFT study, which can be seen in upto 25% of patients (28). The effective dose-equivalent for a99mTc-WBC study is approximately 3 mSv, whereas it is onthe order of 6 mSv for a barium SBFT or 8.5 mSv for abarium enema (29). The99mTc-WBC scan at our institutioncosts approximately $1000 and a colonoscopy with ileos-copy with biopsy costs approximately $2400 (when profes-sional fee, procedure room fee, pathology interpretation fee,and histology fee are all considered).

The presence of a GI bleed occurring at the same time asthe 99mTc-WBC study can complicate the interpretation offindings. Scintigraphy with99mTc-WBC is not useful indefining anatomic details such as fibrotic strictures, prest-enotic dilations, or fistulas, which are best evaluated bybarium radiographic studies. Conversely, these complica-tions of IBD are unlikely to be seen at first presentationwhen the child is screened with99mTc-WBC scintigraphy.Some have reported difficulties assessing inflammation inthe rectum using99mTc-WBC (Fig. 2). The difficulty stemsfrom the fact that uptake in the bladder can overlap therectum (18, 30–33). SPECT images and/or 3-D imaging cancircumvent this limitation (1). Three-dimensional movies of99mTc-WBC can be viewed at this web page: HTTP://www.arad.upmc.edu/users/charron/index.htm.

In conclusion, our findings indicate that99mTc-WBCscintigraphy is useful to confirm or rule out inflammation inchildren suspected of having IBD. A negative99mTc-WBCscan in a patient with bowel symptoms virtually excludesactive IBD. Conversely, a positive99mTc-WBC scan ishighly suggestive of a diagnosis of IBD. When the99mTc-WBC scan is positive the study should be followed bycolonoscopy to further characterize the histopathologicalnature of the colitis. There appears to be a diminished rolefor conventional radiology in the initial investigation ofIBD. Our data suggest that the UGI-SBFT studies have alow sensitivity and should probably not be used initially ina child suspected of having IBD.

Reprint requests and correspondence:Martin Charron, M.D.,F.R.C.P.(C), Division of Nuclear Medicine, Department of Radi-ology, Children’s Hospital of Philadelphia, 34th Street and CivicCenter Boulevard, Philadelphia, PA 19104.

Received Mar. 26, 1999; accepted Oct. 26, 1999.

REFERENCES

1. Charron, M, Del Rosario JF, Kocoshis S. Pediatric inflamma-tory bowel disease: Assessment with scintigraphy with99mTcwhite blood cells. Radiology 1999;212:507–13.

2. Gibson P, Lichtenstein M, Salehi N, et al. Value of positivetechnetium-99mleukocyte scans in predicting intestinal inflam-mation. Gut 1991;32:1502–7.

3. Charron, M, Del Rosario JF, Kocoshis S. Assessment of ter-minal ileal and colonic inflammation in Crohn’s disease withTc99m-WBC. Acta Paediatrica 1999;88:193–8.

4. Charron, M, Del Rosario JF, Kocoshis S. Comparison of thesensitivity of early versus delayed imaging with Tc-99m HM-PAO WBC in children with inflammatory bowel disease (inprocess citation). Clin Nucl Med 1998;23:649–53.

5. Burbige EJ, Huang SH, Bayless TM. Clinical manifestationsof Crohn’s disease in children and adolescents. Pediatrics1975;55:866–71.

6. Kirschner BS. Ulcerative colitis and Crohn’s disease in chil-dren. Diagnosis and management. Gastroenterol Clin NorthAm 1995;24:99–117.

7. Gomes P, du Boulay C, Smith CL, et al. Relationship betweendisease activity indices and colonoscopic findings in patientswith colonic inflammatory bowel disease. Gut 1986;27:92–5.

8. Landi B, Anh TN, Cortot A, et al. Endoscopic monitoring ofCrohn’s disease treatment: A prospective, randomized clinicaltrial. The Groupe d’Etudes Therapeutiques des AffectionsInflammatoires Digestives. Gastroenterology 1992;102:1647–53.

9. Pera A, Bellando P, Caldera D, et al. Colonoscopy in inflam-matory bowel disease. Diagnostic accuracy and proposal of anendoscopic score. Gastroenterology 1987;92:181–5.

10. Quinn PG, Binion DG, Connors PJ. The role of endoscopy ininflammatory bowel disease. Med Clin North Am 1994;78:1331–52.

11. Cellier C, Sahmoud T, Froguel E, et al. Correlations betweenclinical activity, endoscopic severity, and biological parame-ters in colonic or ileocolonic Crohn’s disease. A prospectivemulticentre study of 121 cases. The Groupe d’Etudes Thera-peutiques des Affections Inflammatoires Digestives. Gut1994;35:231–5.

12. Arndt JW, Grootscholten MI, van Hogezand RA, et al. In-flammatory bowel disease activity assessment using techne-tium-99m-HMPAO leukocytes. Dig Dis Sci 1997;42:387–93.

13. Giaffer MH, Tindale WB, Holdsworth D. Value of techne-tium-99mHMPAO-labelled leucocyte scintigraphy as an initialscreening test in patients suspected of having inflammatorybowel disease. Eur J Gastroenterol Hepatol 1996;8:1195–200.

14. Celentano L, Cirillo LC, D’Arienzo A, et al. Diagnostic ac-curacy, and clinical significance of combined small bowelenema, and scintigraphy with pure fractions of neutrophilgranulocytes labeled with HMPAO-Tc99min chronic intestinalinflammation. Radiol Med (Torino) 1996;92:398–404.

15. Allan RA, Sladen GE, Bassingham S, et al. Comparison ofsimultaneous99mTc-HMPAO and 111 in oxine-labelled whitecell scans in the assessment of inflammatory bowel disease.Eur J Nucl Med 1993;20:195–200.

16. Becker W, Fischbach W, Weppler M, et al. Radiolabelledgranulocytes in inflammatory bowel disease: Diagnostic pos-

1211AJG – May, 2000 99mTc Leukocyte Scintigraphy and SBFT in Children

sibilities and clinical indications. Nucl Med Commun 1988;9:693–701.

17. Dhote R, Beades E, Le Dinh T, et al. Scintigraphy usingleukocytes marked with99m-Tc-HMPAO in Crohn’s disease.Acta Gastroenterol Belg 1995;58:353–63.

18. Kennan N, Hayward M. Tc HMPAO-labelled white cell scin-tigraphy in Crohn’s disease of the small bowel. Clin Radiol1992;45:331–4.

19. Mairal L, de Lima PA, Martin-Comin J, et al. Simultaneousadministration of 111In-human immunoglobulin and99mTc-HMPAO labelled leucocytes in inflammatory bowel disease.Eur J Nucl Med 1995;22:664–70.

20. Malcolm PN, Bearcroft CP, Pratt PG, et al. Technetium-99m

hexamethylpropyleneamineoxime labelled leucocyte scanningin the initial diagnosis of Crohn’s disease (see comments). Br JRadiol 1994;67:964–8.

21. Jobling JC, Lindley KJ, Yousef Y, et al. Investigating inflam-matory bowel disease—White cell scanning, radiology, andcolonoscopy. Arch Dis Child 1996;74:22–6.

22. Jewell FM, Davies A, Sandhu B, et al. Technetium-99m-HM-PAO labelled leucocytes in the detection and monitoring ofinflammatory bowel disease in children. Br J Radiol 1996;69:508–14.

23. Papos M, Varkonyi A, Lang J, et al. HM-PAO-labeled leuko-cyte scintigraphy in pediatric patients with inflammatorybowel disease. J Pediatr Gastroenterol Nutr 1996;23:547–52.

24. Charron M, Fernando del Rosario J, Kocoshis S. Distributionof acute bowel inflammation determined by technetium-la-beled white blood cells in children with inflammatory boweldisease. Inflamm Bowel Dis 1998;4:84–8.

25. Prantera C, Luzi C, Olivotto P, et al. Relationship betweenclinical and laboratory parameters and length of lesion inCrohn’s disease of small bowel. Dig Dis Sci 1984;29:1093–7.

26. Meuwissen SG, Pape KS, Agenant D, et al. Crohn’s disease ofthe colon. Analysis of the diagnostic value of radiology, en-doscopy, and histology. Am J Dig Dis 1976;21:81–8.

27. Charron M, del Rosario JF, Kocoshis S. Use of technetium-tagged white blood cells in patients with Crohn’s disease andulcerative colitis: Is differential diagnosis possible? PediatrRadiol 1998;28:871–7.

28. Coremans G, Rutgeerts P, Geboes K, et al. The value ofileoscopy with biopsy in the diagnosis of intestinal Crohn’sdisease. Gastrointest Endosc 1984;30:167–72.

29. Shields RA, Lawson RS. Effective dose equivalent. Nucl MedCommun 1987;8:851–5.

30. Lantto E, Jarvi K, Krekela I, et al. Technetium-99mhexamethylpropylene amine oxine leucocytes in the assessment of diseaseactivity in inflammatory bowel disease. Eur J Nucl Med 1992;19:14–8.

31. Sciarretta G, Furno A, Mazzoni M, et al. Technetium-99m

hexamethyl propylene amine oxime granulocyte scintigraphyin Crohn’s disease: Diagnostic and clinical relevance (seecomments). Gut 1993;34:1364–9.

32. Peters AM, Roddie ME, Danpure HJ, et al. 99Tcm-HMPAOlabelled leucocytes. Comparison with 111In-tropolonate la-belled granulocytes. Nucl Med Commun 1988;9:449–63.

33. Kolkman JJ, Falke TH, Roos JC, et al. Computed tomographyand granulocyte scintigraphy in active inflammatory boweldisease. Comparison with endoscopy and operative findings.Dig Dis Sci 1996;41:641–50.

1212 Charron et al. AJG – Vol. 95, No. 5, 2000