- Email: [email protected]
Location and Activity of Ulcerative and Crohn's Colitis by Indium 111 Leukocyte Scan
GASTROENTEROLOGY 1983;84:388-93
Location and Activity of Ulcerative and Crohn's Colitis by Indium 111 Leukocyte Scan A Prospective Comparison Study DAVID T. STEIN, GARY M. GRAY, PETER B. GREGORY, MALCOLM ANDERSON, DAVID A. GOODWIN, and 1. ROSS McDOUGALL Divisions of Gastroenterology, Diagnostic Radiology, and Nuclear Medicine, Stanford University School of Medicine, Stanford, California
A prospective blinded study comparing the indium 111 leukocyte scan to barium enema, colonoscopy, or surgery or a combination of these, was carried out in 15 patients (10 with active ulcerative colitis and 5 with active Grohn's colitis). Correlation of disease location to colonic regions between indium scan and other diagnostic studies was excellent in 11 instances, good in 2, and poor in 3. In 2 of the 3 studies where major disagreement occurred, the comparative barium enema was performed >2 mo after the indium scan. Disease activity, estimated by the intensity of radionuclide uptake, was compared to clinical disease activity assessed by the Grohn's Disease Activity Index for both forms of colitis. The relative degree of inflammation estimated by the indium scan correlated well with the independent clinical assessment (correlation coefficient = 0.81). The indium 111 leukocyte scan appears to be an accurate, noninvasive method for assessing the extent and the severity of the inflammation in patients with acute ulcerative or Grohn's colitis. The extent of bowel involvement during an acute attack of ulcerative colitis or Grohn's disease cannot be assessed reliably by clinical examination alone (1). More accurate invasive diagnostic studies should usually not be done because they may worsen the patient's condition (2-7). For that reason, several weeks may elapse before invasive studies to determine the extent of the disease can be justified. A simple, noninvasive method to identify the locaReceived December 17, 1981. Accepted September 23, 1982. Address requests for reprints to: Dr. Gary M. Gray, Gastroenterology Division, Stanford University Medical Center, Stanford, California 94305. © 1983 by the American Gastroenterological Association 0016-5085/83/020388-06$03.00
tion and activity of the inflammation would, therefore, be a useful adjunct in the management of acute inflammatory bowel disease. Autologous leukocytes, when exposed to indium 111 and 8-hydroxyquinoline (oxine) chelating agent (8,9), become stably labeled with the radionuclide (10). The leukocytes can then be reinjected and will migrate to areas of focal inflammation (11). This imaging technique has been shown to identify abcesses (12-14) and sites of occult intestinal hemorrhage (15), and has been useful in investigation of granulocyte function (10,16). Indium 111 has emission peaks that are readily detected by commonly available gamma-scanning devices. The labeled leukocytes retain much of their functional viability (10), and can be studied within hours or for 2-3 days after injection (17,18). In this prospective study, we have compared the indium 111 leukocyte scan with other modalities usually used to assess activity and extent of inflammatory bowel disease. This imaging technique appears to provide an accurate noninvasive method for determining the extent and activity of acute ulcerative and Grohn's colitis.
Methods The method of labeling leukocytes has been adapted from Thakur (19). Whole blood (40 ml) was drawn from the patient and added to a tube containing 500 U of preservative-free heparin, 3 ml hydroxyethylstarch (Hetastarch, McGaw Laboratories Irvine, Calif.), and 3 ml physiological nonbacteriostatic saline. The mixture was left in a vertical position for 1 h to allow erythrocytes and some lymphocytes to sediment by gravity. The platelet- and granulocyte-rich supernatant was removed and centri-
February 1983
INDIUM 111 LEUKOCYTE SCAN IN COLITIS
389
A Figure 1. A. Anterior whole body indium 111 leukocyte scan 24 h after injection of labeled cells in a pati ent with acute ulcerative colitis. Normal distribution of labeled cells is seen in liver, spleen, and bone marrow. Whereas no colonic uptake of labeled leukocytes normally occurs, uptake is present throughout th e entire colon in this patient. A fo cus of uptake in the right thorax represents an infected central venous catheter that was fi rst detected by this scan. B. Air contrast barium enema in the same patient demonstrating generalized mucosal disease manifested by loss of haustrations, ulcerations . and pseudo pol yps most prominent in the descend ing and sigmoid colon.
fuged at 250 g for 5 min to form a leukocyte pellet consisting almost entirely of polymorphonuclear cells. The platelet-rich supernatant was removed and spun at 1000 g for 30 min to yield platelet-free plasma. The whitecell pellet was resuspended in 5 ml saline and 1 mCi of indium 111 oxine (Diagnostic Isotopes Bloomfield, N.J. or Mediphysics, Emeryville, CaliL) in 50 J.LI ethanol; 150 J.LI saline was added slowly while swirling the white-cell suspension. The final mixture was then incubated for 30 min at room temperature and centrifuged at 250 g to form a pellet of labeled leukocytes. The efficiency of cell labeling was calculated by measuring the radioactivity of the final supernatant and leukocyte pellet. Seventy-five to ninetyfive percent of radioactivity was found in the cell pellet. The leukocytes were resuspended in 5-10 ml of plateletfree plasma and returned to the patient by intravenous injection. Scans were obtained 18-24 h after reinjection of labeled cells using either a Picker Dyna-4 camera with moving table (Picker Corp ., Highland Heights, Ohio) or a Searle tomographic scanner (Searle Pharmaceuticals, Inc. , Chicago, Ill.) . Anterior, posterior, and lateral views (or tomographic scanning) were obtained in all patients to minimize potential interference by hepatic and splenic uptake of radiolabeled cells. A typical example of a positive white-cell scan in a patient with acute ulcerative colitis is shown in Figure 1. In comparison to other diagnostic techniques, the cost of the indium scan at our
institution is similar to air contrast barium enema and gallium 67 scan, and approximately one-half the cost of co lonoscopy.
Patients Patients with active colitis were enrolled prospectively from the Stanford University and Palo Alto Veterans Administration Medical Centers between April 1978 and April 1981. The study was approved by the Human Subjects Committee at both institutions. All patients had active ulcerative or Crohn 's colitis on the basis of the clinical history, physical examination, and sigmoidoscopic appearance (1,20). The absence of ova, parasites, or pathogens was confirmed by stool examination and culture. Some patients had a history of chronic ulcerative or Crohn's colitis and presented with recurrent symptoms. Written informed consent was obtained, and the clinical severity and activity of disease was judged and recorded by the study physician. Patients were managed as seen fit by the primary nonstudy physicians responsible for their care. Whenever corticosteroids or sulfasalazine were being administered, they were continued at the same dosage on the day of the indium scan. The radio labeled leukocyte scan was obtained as soon as possible after admission to the study. Barium enema, colonoscopy, or both, were completed when the patient had recovered sufficiently to
390
STEIN ET AL.
allow the test to be accomplished without undue risk; surgery was performed as needed for proper patient care.
Location of Disease Indium Scan vs. X-Ray, Colonoscopy, and Surgery Two nuclear medicine physicians (1. R. McD., D. A. G.) independently interpreted the indium 111 scans with no knowledge of the patient's identity, condition, or other diagnostic studies. Radionuclide activity was recorded as present or absent in eight separate segments of colon corresponding to cecum, ascending colon, hepatic flexure, transverse colon, splenic flexure, descending colon, sigmoid, and rectum. Barium enema examinations (11 double contrast, 2 single contrast) were interpreted in a similar fashion by a radiologist (M. F. A.) and a gastroenterologist (G. M. G.). Each of the four readers assessed each study twice to determine internal consistency. The two nuclear medicine physicians had an interobserver agreement of 94% and an intraobserver agreement of 89% in reading the indium scans. The barium enema interobserver and intraobserver agreement was 69% and 96%, respectively. Intraobserver differences in interpretation were subjected to a blinded third reading and interobserver discrepancies in determining disease location were resolved by a joint decision between observers. Disease location and activity at colonoscopy were judged by physicians unaware that the study was in progress. Surgical assessment of disease was obtained from the resected colon or from intraoperative visualization of disease and preoperative sigmoidoscopy.
Activity of Disease Indium Scan vs. Clinical Assessment Clinical disease activity for both forms of colitis was graded using the Crohn's Disease Activity Index (CD AI) (21). The activity score was recorded before obtaining the indium scan, and results of the scan were not released to the clinical study physicians. The CDAI scores have been converted to descriptive categories of disease activity by the National Crohn's Cooperative Disease Study Group (22) (mild: <175; moderate: 175-300; moderately severe: 301-450; severe: >450). The activity of disease was interpreted on the indium scan by comparing the intensity of radionuclide uptake in the eight segments of colon to the intensity of uptake by bone marrow, liver, and spleen (23). Disease activity was judged as mild if the intensity of radionuclide uptake by the diseased colon equaled uptake by the bone marrow; moderate if it equaled uptake by the liver; moderately severe if it equaled uptake by the spleen, and severe if it exceeded uptake by the spleen. Scans were graded in an independent blinded manner by two observers (1. R. McD.; D. A. G.) with no knowledge of the patient's clinical status or other diagnostic studies.
GASTROENTEROLOGY Vol. 84, No.2
Results Study Patients Fifteen patients (10 with ulcerative and 5 with Crohn's colitis) were admitted to the study (Table 1). Nine were men and 6 were women. The average age at the time of entry was 35 yr, with a range of 16-63 yr. The mean prestudy duration of disease was 39 mo, with a range of 0-21 yr. Six patients presented with their first attack of colitis. Site of colon involvement. The location of radionuclide uptake on indium 111 leukocyte scan was compared to location of colitis as estimated by barium enema (13 patients), colonoscopy (3 patients), surgery (2 patients), or a combination of these (Table 1). A typical example of this correlation is given in Figure 1. The median time interval between scan and radiograph was 43.5 days. Seventeen indium scans were completed in 15 patients. The correlation between indium scan and other diagnostic studies was judged perfect (no discrepancy for each of the eight colon segments) or good (discrepancy for one colonic segment) in 11 instances (69%) (Table 2). One additional scan could not be used to assess location of disease because the patient was too ill to undergo an invasive study until many weeks later. When symptoms did improve, a repeat scan and a barium enema radiograph were obtained (case 5). In 2 of the 3 studies where there was poor correlation between white blood cell scan and radiograph, the time interval between studies was >80 days. Only 1 patient, where the correlation was moderate or better, had an interval between studies >80 days. Assessment of clinical activity. The Crohn's Disease Activity Index was used to quantify clinical disease activity for both types of colitis. The mean CD AI score in our population was 359, with a range of 0-601. The median interval between the initial clinical examination and the ll1indium scan was 4.5 days with a range of 1-8 days. One patient's (case 7) CD AI score could not be accurately assessed for comparison to his actual clinical activity because, being mentally retarded and quadriplegic, he could not accurately answer the necessary CDAI questions and had very little abdominal sensation. We have eliminated this case from the activity correlation analysis (see Discussion), but have retained it for estimation of disease location. That he had severe colitis was documented by the development of toxic megacolon requiring colostomy. The correlation between the clinical grade of disease activity and the activity estimated from the leukocyte scan was perfect (no discrepancy) in 76% (13 of 17) and very good (discrepancy by one category of activity) in 24% (4 of 17). The correlation
INDIUM 111 LEUKOCYTE SCAN IN COLITIS
February 1983
391
Table 1. Location and Activity of Inflammatory Bowel Disease Patient number 1 2a c
2b 3 4 5aC 5b 6 7 8 9 10
tl
12 13 14 15 o
Location of disease b Other studies
CDAI score
Indium scan
C A H T S D Sg R
C A HT S D Sg R
+
+
(Category) 194 (M) 350 (MS) (M) 153 (M) 234 (M) 291 (S) 453 (M) 0 367 (MS) - d (S) 389 (MS) 428 (MS) 327 (Mod) (S) 593 401 (MS) (S) 601 382 (MS) 277 (Mod)
MS MS none M Mod S none MS S MS MS Mod S S MS MS Mod
+ + + + + + + + + + + + + + - - - - - - -
CC UC UC UC UC UC UC CC UC UC CC UC CC UC UC UC CC
Activity of disease
Indium scan
Diagnosis O
+ - + + + - - + + + + + + + + + + + + + + + +
+ + + +
-
-
+ + + + -
-
-
-
-
+ + + +
+ + +
+ + - - - - + + + + + + + - - - - - - - + - - - + + + + + + + not done
- + + + + + + + + + + + + + + + + + + + - - + + + - - - + + + + + + - - + + + + + + + + + -
-
+ + + +
+ + + +
+ + + + + + + -
+ + + + +
+ + + + + + + + +
+ + + + + +
-
-
+ +
-
+ + + +
+ + + - + + - + -
+ + + + + + + +
CC = Crohn's colitis; UC = ulcerative colitis. b CAHTSDSgR indicates colonic segments from cecum (left) to rectum (right) (C = cecum; A = ascending; H = hepatic flexure; T = transverse; S = splenic flexure; D = descending; Sg = sigmoid; R = rectum); (+) denotes involvement and (-) absence of involvement of the particular segment. M = mild; Mod = moderate; MS = moderately severe; S = severe; None = no uptake. C Letters indicate study before (a) and after (b) therapy for acute colitis. d CDAI score could not be estimated (see text).
between the clinical assessment (CDAI score) and the indium scan grading (r = 0.81) is shown in Figure 2. In no instance did the scan fail to show radionuclide uptake when the colitis was clinically active. Two patients (cases 2 and 5) had indium scans during acute colitis and repeat scans when asymptomatic. In both cases, radionuclide uptake in the colon changed from universal uptake to no uptake whatsoever. Eleven of 17 scans (65%) were completed on patients who were taking some form of corticosteroid. The dosage of oral prednisone (average 45 mg) or intravenous methylprednisolone (averge 65 mg), although relatively high, did not appear to inhibit leukocyte migration to the inflamed colon. Although the leukocyte scan appeared to readily Table 2. Correlation of Indium Scan and Other Studies for Disease Location Agreement category Perfect Good Moderate Poor
No. of discrepant colon segments O None 1 2-3 3
No. of scans in categoryb 8 3 2 3
(50.5%) (18.5%) (12.5%) (18.5%)
°Number of colon segments in disagreement between scan and other studies (Barium enema x-rays; colonoscopy; surgical specimen). b Sixteen scans in 15 patients; patient 2 studied twice and patient 5a not included (See text and Table 1.)
identify colonic inflammation, it did not reliably differentiate ulcerative and Crohn's colitis. The diagnosis was usually apparent, however, after clinical assessment and sigmoidoscopic examination.
Discussion The indium 111 leukocyte scan has several unique characteristics that favor its use as a diagnostic modality in acute ulcerative and Crohn's colitis. The test is noninvasive, easily tolerated, simple to perform, and utilizes commonly available gammascanning equipment. The radionuclide hC!s emission peaks and a physical half-life that make it ideal for clinical use (17,24), and indium lll-labeled leukocytes show no uptake in the normal gastrointestinal tract (14,25). These qualities allow its use during active or recurrent colitis and in patients who are too fragile for invasive studies. All of the patients in this study had abnormal radionuclide uptake in the colon during episodes of active colitis. The scan findings correlated well with those obtained by established techniques. Where the correlation was poor, there was usually a lengthy delay between the white blood cell scan and the subsequent radiographic examination. Other explanations for discrepant findings include the tendency for labeled leukocytes shed into the bowel lumen to migrate with the fecal stream to areas where no disease is present, and recognition that the scan may
392
STEIN ET AL.
delineate areas of inflammation not apparent on barium enema. Direct observation of the mucosa by colonoscopy will often detect such areas. It should also be recognized that labeled leukocytes will not migrate to regions of inactive disease whereas these regions might be readily apparent with other diagnostic techniques. There is no established grading system for clinical disease activity that can be applied to both ulcerative and Crohn's colitis. The CDAI (21,22) accurately grades Crohn's disease, but has not been used in ulcerative colitis. The parameters that are measured to obtain the CDAI score, however, are also major indicators of disease activity in ulcerative colitis (sense of well-being, abdominal pain, number of watery or bloody stools per day, constitutional symptoms, and extracolonic manifestations of disease). Since we felt it important to have a common grading system for both diseases, the CD AI was chosen because it is an established system in Crohn's disease and because it correlated well with the physician's subjective assessment of activity in ulcerative colitis. In only 1 case did the CDAI misrepresent disease activity; that occurred in a mentally retarded quadriplegic patient who was unable to respond to the appropriate CDAI questions (case 7). The indium 111 leukocyte scan identified active colitis, whether Crohn's or ulcerative, and whether new or recurrent in onset. In addition, the use of corticosteroids did not appear to inhibit the migration of labeled cells to the diseased colon. There was a remarkably good correlation between the clinical assessment of disease activity and the degree of radionuclide uptake. This suggests that the indium scan may be useful in following the course of disease activity. As anticipated, the leukocyte scan did not distinguish between the two types of colitis. Patchy or right-sided uptake was evident in some cases of Crohn's disease, but no specific diagnosis could be made in most patients with universal colitis. Yet, the indium scan adds a diagnostic dimension to sigmoidoscopy, barium enema, and colonoscopy. If a diagnosis can be made through clinical assessment and sigmoidoscopy, then the scan offers an easy noninvasive method to define the extent and activity of disease. No significant complications were encountered during the use of this diagnostic technique. No colon preparation was necessary before scanning, and all of our patients tolerated the procedure easily. The severity of colitis was not intensified by the study and we observed no cases of megacolon, perforation, sepsis, or hemorrhage that were temporally related to radionuclide imaging. The total whole body radiation exposure with the
GASTROENTEROLOGY Vol. 84. No.2
600
•
x w o z
>-
~
:> ~
u <[ ....J
<[
U Z
....J U
o~------~------~------~------~
o
2
III
3
4
In - SCAN SCORE
Figure 2. Correlation of clinical estimation of disease activity (CDAI. see Methods) and indium scan interpretation of activity (1 = mild. 2 = moderate. 3 = moderately severe. and 4 = severe).
indium 111 leukocyte scan is approximately 0.31 rad/500 JLCi dose, with the highest concentrations seen in the liver (1.3 rads/500 JLCi) , spleen (8.5 rads/500 JLCi), and bone marrow (2.3 rads/500 JLCi) (22). A barium enema with fluoroscopy and spot films gives a slightly greater whole body dose, but less intense exposure to the liver, spleen, and bone marrow (26). A standard gallium 67 scan (5 mCi dose) gives comparable exposure to these organs. Although previous pr~liminary reports have shown positive indium scans in inflammatory bowel disease (27), our study extends those findings by identifying patients prospectively and comparing the scan results with an assessment of clinical status and with the findings on barium enema or colonoscopy. This diagnostic technique may be of value in defining the location of colitis early in the course of acute disease when barium contrast studies are contraindicated. Its eventual clinical application will depend on future experience with larger numbers of patients in a variety of clinical situations. Although its role in the management of inflammatory bowel disease remains to be established, the indium scan
February 1983
appears to have great potential in assessing the activity and extent of Crohn's and ulcerative coliti~
INDIUM 111 LEUKOCYTE SCAN IN COLITIS
15.
References 1. Edwards FC, Truelove C. The course and prognosis of ulcerative colitis. Gut 1963;4:299-315. 2. Wruble LD, Bronstein MW. Toxic dilatation of the colon following barium enema examination during the quiescent stage of chronic ulcerative colitis. Am J Dig Dis 1968;13:91823. 3. Binder SC, Patterson JF, Glotzer DJ. Toxic megacolon in ulcerative colitis. Gastroenterology 1974;66:909-15. 4. Jalan KN, Sircul W, Card WI, et al. An experience of ulcerative colitis. l. Toxic dilatation in 55 cases. Gastroenterology 1969;57:68-82. 5. Norland CC, Kirsner JB. Toxic dilatation of colon (toxic megacolon): etiology, treatment and prognosis in 42 patients. Medicine 1969;48:229-50. 6. Waye JD. The role of colonoscopy in the differential diagnosis of inflammatory bowel disease. Gastrointest Endosc 1977;23:150-4. 7. Myren J. Eie H, Serck-Hanssen A. The diagnosis of colitis by colonoscopy, with biopsy and x-ray examination. A blind comparative study. Scand J GastroenteroI1976;11:141-4. 8. Thakur ML, Coleman RE, Welch MJ. Indium-lll-labeled leukocytes for the localization of abscesses: preparation, analysis. tissue distribution and comparison with gallium-67 citrate in dogs. J Clin Lab Med 1977;89:217-28. 9. Thakur ML, Coleman RE, Mayhall CG, Welch MJ Jr. Preparation and evaluation of 111 In-labeled leukocytes as an abscess imaging agent in dogs. Radiology 1976;119:731-2. 10. Zakhireh B, Thakur ML, Malech HL, Cohen MS, Gottschalk A. Root RK. Indium-l11-labeled human polymorphonuclear leukocytes: viability, random migration, chemotaxis, bactericidal capacity, and ultrastructure. J Nucl Med 1979;20:741-7. 11. Thakur ML, Lavender JP, Arnot RN, Silvester DJ, Segal AW. Indium-lll-labeled autologous leukocytes in man. J Nucl Med 1977;18:1014-21. 12. Segal AW, Thakur ML, Arnot RN, Lavender JP. Indium-l11labeled leukocytes for localization of abscesses. Lancet 1976;ii:l056-8. 13. McDougall IR. Baumert JE, Lantieri RL. Evaluation of l11In leukocyte whole body scanning in 59 patients. Am J Roentegenol 1979;133:849-54. 14. Goodwin DA, Doherty PW, McDougall IR. Clinical use of indium-l11 labeled white cells. An analysis of 312 cases of
16. 17. 18.
19. 20.
21.
22. 23.
24.
25. 26.
27.
393
radiolabeled blood cellular elements. In: Gottschalk A, Thakur ML, eds. New York: Trivirum Publishers, 1981:131-45. Winzelberg GG, Castronovo FP, Callahan RJ, McKusick KA, Strauss HW. 111In oxine labeled red cells for detection of simulated lower gastrointestinal bleeding in an animal model. Radiology 1980;135:455-61. Weibel BJ, Forstrom L, McCullough J. Studies of the kinetics of indium-l11-labeled granulocytes. J Lab Clin Med 1979;94:246-55. Dutcher JP, Schiffer CA, Johnston GS. Rapid migration of l11indium-labeled granulocytes to sites of infections. N Engl J Med 1981;304:586-9. Ferrant A, Dehasque N, Leners N, Neunier H. Scintigraphy with In-l11-labeled red cells in intermittent gastrointestinal bleeding. J Nucl Med 1980;21:844-5. Thakur ML. Indium-lll: a new radioactive tracer for leukocytes. Exp Hematol 1977;5(suppl):145-50. Sparberg M, Fennessy J, Kirsner JB. Ulcerative proctitis and mild ulcerative colitis: a study of 220 patients. Medicine 1966;45:391. Best WR, Bectel JM, Singleton JW, Kern F Jr. Development of a Crohn's disease activity index. National Cooperative Crohn's Disease Study. Gastroenterology 1976;70:439-44. Best WR, Bectel JM, Singleton JW. Rederived values of the eight coefficients of the Crohn's disease activity index (CDAI). Gastroenterology 1979;77:843-6. Goodwin DA, Finston RA, Smith Sl. The distribution and dosimetry of In-111-labeled leukocytes and platelets in humans. Proceedings of the Third International Radiopharmaceutical Dosimetry Symposium. Oak Ridge, Tenn: Food and Drug Administration, 1980:88-101. Goodwin DA, Sundberg MW, Diamanti CI, Meares CF. l11In_ labeled radiopharmaceuticals and their clinical use. In: Subramanian GS, Rhodes BA, Cooper JF, and Sodd VJ, eds. Radiopharmaceuticals. New York: Society of Nuclear Medicine, 1975:80-101. Saverymuttu SH, Peters AM, Lavender JP, Hodgson HJ, Chadwick VS. Indium 111 labelled autologous leukocytes in inflammatory bowel disease (abstr). Gastroenterology 1981;80:1273. Shleien B, Tucker TT, Johnson DW. The mean active bone marrow dose to the adult population of the United States from diagnostic radiology. Washington, D.C.: U.S. Department of Health, Education, and Welfare, Bureau of Radiological Health, January 1977. Segal AW, Munro JM, Ensell J, Sarner M. Indium-lll tagged leukocytes in the diagnosis of inflammatory bowel disease. Lancet 1981 ;ii:230-2.
Location and Activity of Ulcerative and Crohn's Colitis by Indium 111 Leukocyte Scan A Prospective Comparison Study DAVID T. STEIN, GARY M. GRAY, PETER B. GREGORY, MALCOLM ANDERSON, DAVID A. GOODWIN, and 1. ROSS McDOUGALL Divisions of Gastroenterology, Diagnostic Radiology, and Nuclear Medicine, Stanford University School of Medicine, Stanford, California
A prospective blinded study comparing the indium 111 leukocyte scan to barium enema, colonoscopy, or surgery or a combination of these, was carried out in 15 patients (10 with active ulcerative colitis and 5 with active Grohn's colitis). Correlation of disease location to colonic regions between indium scan and other diagnostic studies was excellent in 11 instances, good in 2, and poor in 3. In 2 of the 3 studies where major disagreement occurred, the comparative barium enema was performed >2 mo after the indium scan. Disease activity, estimated by the intensity of radionuclide uptake, was compared to clinical disease activity assessed by the Grohn's Disease Activity Index for both forms of colitis. The relative degree of inflammation estimated by the indium scan correlated well with the independent clinical assessment (correlation coefficient = 0.81). The indium 111 leukocyte scan appears to be an accurate, noninvasive method for assessing the extent and the severity of the inflammation in patients with acute ulcerative or Grohn's colitis. The extent of bowel involvement during an acute attack of ulcerative colitis or Grohn's disease cannot be assessed reliably by clinical examination alone (1). More accurate invasive diagnostic studies should usually not be done because they may worsen the patient's condition (2-7). For that reason, several weeks may elapse before invasive studies to determine the extent of the disease can be justified. A simple, noninvasive method to identify the locaReceived December 17, 1981. Accepted September 23, 1982. Address requests for reprints to: Dr. Gary M. Gray, Gastroenterology Division, Stanford University Medical Center, Stanford, California 94305. © 1983 by the American Gastroenterological Association 0016-5085/83/020388-06$03.00
tion and activity of the inflammation would, therefore, be a useful adjunct in the management of acute inflammatory bowel disease. Autologous leukocytes, when exposed to indium 111 and 8-hydroxyquinoline (oxine) chelating agent (8,9), become stably labeled with the radionuclide (10). The leukocytes can then be reinjected and will migrate to areas of focal inflammation (11). This imaging technique has been shown to identify abcesses (12-14) and sites of occult intestinal hemorrhage (15), and has been useful in investigation of granulocyte function (10,16). Indium 111 has emission peaks that are readily detected by commonly available gamma-scanning devices. The labeled leukocytes retain much of their functional viability (10), and can be studied within hours or for 2-3 days after injection (17,18). In this prospective study, we have compared the indium 111 leukocyte scan with other modalities usually used to assess activity and extent of inflammatory bowel disease. This imaging technique appears to provide an accurate noninvasive method for determining the extent and activity of acute ulcerative and Grohn's colitis.
Methods The method of labeling leukocytes has been adapted from Thakur (19). Whole blood (40 ml) was drawn from the patient and added to a tube containing 500 U of preservative-free heparin, 3 ml hydroxyethylstarch (Hetastarch, McGaw Laboratories Irvine, Calif.), and 3 ml physiological nonbacteriostatic saline. The mixture was left in a vertical position for 1 h to allow erythrocytes and some lymphocytes to sediment by gravity. The platelet- and granulocyte-rich supernatant was removed and centri-
February 1983
INDIUM 111 LEUKOCYTE SCAN IN COLITIS
389
A Figure 1. A. Anterior whole body indium 111 leukocyte scan 24 h after injection of labeled cells in a pati ent with acute ulcerative colitis. Normal distribution of labeled cells is seen in liver, spleen, and bone marrow. Whereas no colonic uptake of labeled leukocytes normally occurs, uptake is present throughout th e entire colon in this patient. A fo cus of uptake in the right thorax represents an infected central venous catheter that was fi rst detected by this scan. B. Air contrast barium enema in the same patient demonstrating generalized mucosal disease manifested by loss of haustrations, ulcerations . and pseudo pol yps most prominent in the descend ing and sigmoid colon.
fuged at 250 g for 5 min to form a leukocyte pellet consisting almost entirely of polymorphonuclear cells. The platelet-rich supernatant was removed and spun at 1000 g for 30 min to yield platelet-free plasma. The whitecell pellet was resuspended in 5 ml saline and 1 mCi of indium 111 oxine (Diagnostic Isotopes Bloomfield, N.J. or Mediphysics, Emeryville, CaliL) in 50 J.LI ethanol; 150 J.LI saline was added slowly while swirling the white-cell suspension. The final mixture was then incubated for 30 min at room temperature and centrifuged at 250 g to form a pellet of labeled leukocytes. The efficiency of cell labeling was calculated by measuring the radioactivity of the final supernatant and leukocyte pellet. Seventy-five to ninetyfive percent of radioactivity was found in the cell pellet. The leukocytes were resuspended in 5-10 ml of plateletfree plasma and returned to the patient by intravenous injection. Scans were obtained 18-24 h after reinjection of labeled cells using either a Picker Dyna-4 camera with moving table (Picker Corp ., Highland Heights, Ohio) or a Searle tomographic scanner (Searle Pharmaceuticals, Inc. , Chicago, Ill.) . Anterior, posterior, and lateral views (or tomographic scanning) were obtained in all patients to minimize potential interference by hepatic and splenic uptake of radiolabeled cells. A typical example of a positive white-cell scan in a patient with acute ulcerative colitis is shown in Figure 1. In comparison to other diagnostic techniques, the cost of the indium scan at our
institution is similar to air contrast barium enema and gallium 67 scan, and approximately one-half the cost of co lonoscopy.
Patients Patients with active colitis were enrolled prospectively from the Stanford University and Palo Alto Veterans Administration Medical Centers between April 1978 and April 1981. The study was approved by the Human Subjects Committee at both institutions. All patients had active ulcerative or Crohn 's colitis on the basis of the clinical history, physical examination, and sigmoidoscopic appearance (1,20). The absence of ova, parasites, or pathogens was confirmed by stool examination and culture. Some patients had a history of chronic ulcerative or Crohn's colitis and presented with recurrent symptoms. Written informed consent was obtained, and the clinical severity and activity of disease was judged and recorded by the study physician. Patients were managed as seen fit by the primary nonstudy physicians responsible for their care. Whenever corticosteroids or sulfasalazine were being administered, they were continued at the same dosage on the day of the indium scan. The radio labeled leukocyte scan was obtained as soon as possible after admission to the study. Barium enema, colonoscopy, or both, were completed when the patient had recovered sufficiently to
390
STEIN ET AL.
allow the test to be accomplished without undue risk; surgery was performed as needed for proper patient care.
Location of Disease Indium Scan vs. X-Ray, Colonoscopy, and Surgery Two nuclear medicine physicians (1. R. McD., D. A. G.) independently interpreted the indium 111 scans with no knowledge of the patient's identity, condition, or other diagnostic studies. Radionuclide activity was recorded as present or absent in eight separate segments of colon corresponding to cecum, ascending colon, hepatic flexure, transverse colon, splenic flexure, descending colon, sigmoid, and rectum. Barium enema examinations (11 double contrast, 2 single contrast) were interpreted in a similar fashion by a radiologist (M. F. A.) and a gastroenterologist (G. M. G.). Each of the four readers assessed each study twice to determine internal consistency. The two nuclear medicine physicians had an interobserver agreement of 94% and an intraobserver agreement of 89% in reading the indium scans. The barium enema interobserver and intraobserver agreement was 69% and 96%, respectively. Intraobserver differences in interpretation were subjected to a blinded third reading and interobserver discrepancies in determining disease location were resolved by a joint decision between observers. Disease location and activity at colonoscopy were judged by physicians unaware that the study was in progress. Surgical assessment of disease was obtained from the resected colon or from intraoperative visualization of disease and preoperative sigmoidoscopy.
Activity of Disease Indium Scan vs. Clinical Assessment Clinical disease activity for both forms of colitis was graded using the Crohn's Disease Activity Index (CD AI) (21). The activity score was recorded before obtaining the indium scan, and results of the scan were not released to the clinical study physicians. The CDAI scores have been converted to descriptive categories of disease activity by the National Crohn's Cooperative Disease Study Group (22) (mild: <175; moderate: 175-300; moderately severe: 301-450; severe: >450). The activity of disease was interpreted on the indium scan by comparing the intensity of radionuclide uptake in the eight segments of colon to the intensity of uptake by bone marrow, liver, and spleen (23). Disease activity was judged as mild if the intensity of radionuclide uptake by the diseased colon equaled uptake by the bone marrow; moderate if it equaled uptake by the liver; moderately severe if it equaled uptake by the spleen, and severe if it exceeded uptake by the spleen. Scans were graded in an independent blinded manner by two observers (1. R. McD.; D. A. G.) with no knowledge of the patient's clinical status or other diagnostic studies.
GASTROENTEROLOGY Vol. 84, No.2
Results Study Patients Fifteen patients (10 with ulcerative and 5 with Crohn's colitis) were admitted to the study (Table 1). Nine were men and 6 were women. The average age at the time of entry was 35 yr, with a range of 16-63 yr. The mean prestudy duration of disease was 39 mo, with a range of 0-21 yr. Six patients presented with their first attack of colitis. Site of colon involvement. The location of radionuclide uptake on indium 111 leukocyte scan was compared to location of colitis as estimated by barium enema (13 patients), colonoscopy (3 patients), surgery (2 patients), or a combination of these (Table 1). A typical example of this correlation is given in Figure 1. The median time interval between scan and radiograph was 43.5 days. Seventeen indium scans were completed in 15 patients. The correlation between indium scan and other diagnostic studies was judged perfect (no discrepancy for each of the eight colon segments) or good (discrepancy for one colonic segment) in 11 instances (69%) (Table 2). One additional scan could not be used to assess location of disease because the patient was too ill to undergo an invasive study until many weeks later. When symptoms did improve, a repeat scan and a barium enema radiograph were obtained (case 5). In 2 of the 3 studies where there was poor correlation between white blood cell scan and radiograph, the time interval between studies was >80 days. Only 1 patient, where the correlation was moderate or better, had an interval between studies >80 days. Assessment of clinical activity. The Crohn's Disease Activity Index was used to quantify clinical disease activity for both types of colitis. The mean CD AI score in our population was 359, with a range of 0-601. The median interval between the initial clinical examination and the ll1indium scan was 4.5 days with a range of 1-8 days. One patient's (case 7) CD AI score could not be accurately assessed for comparison to his actual clinical activity because, being mentally retarded and quadriplegic, he could not accurately answer the necessary CDAI questions and had very little abdominal sensation. We have eliminated this case from the activity correlation analysis (see Discussion), but have retained it for estimation of disease location. That he had severe colitis was documented by the development of toxic megacolon requiring colostomy. The correlation between the clinical grade of disease activity and the activity estimated from the leukocyte scan was perfect (no discrepancy) in 76% (13 of 17) and very good (discrepancy by one category of activity) in 24% (4 of 17). The correlation
INDIUM 111 LEUKOCYTE SCAN IN COLITIS
February 1983
391
Table 1. Location and Activity of Inflammatory Bowel Disease Patient number 1 2a c
2b 3 4 5aC 5b 6 7 8 9 10
tl
12 13 14 15 o
Location of disease b Other studies
CDAI score
Indium scan
C A H T S D Sg R
C A HT S D Sg R
+
+
(Category) 194 (M) 350 (MS) (M) 153 (M) 234 (M) 291 (S) 453 (M) 0 367 (MS) - d (S) 389 (MS) 428 (MS) 327 (Mod) (S) 593 401 (MS) (S) 601 382 (MS) 277 (Mod)
MS MS none M Mod S none MS S MS MS Mod S S MS MS Mod
+ + + + + + + + + + + + + + - - - - - - -
CC UC UC UC UC UC UC CC UC UC CC UC CC UC UC UC CC
Activity of disease
Indium scan
Diagnosis O
+ - + + + - - + + + + + + + + + + + + + + + +
+ + + +
-
-
+ + + + -
-
-
-
-
+ + + +
+ + +
+ + - - - - + + + + + + + - - - - - - - + - - - + + + + + + + not done
- + + + + + + + + + + + + + + + + + + + - - + + + - - - + + + + + + - - + + + + + + + + + -
-
+ + + +
+ + + +
+ + + + + + + -
+ + + + +
+ + + + + + + + +
+ + + + + +
-
-
+ +
-
+ + + +
+ + + - + + - + -
+ + + + + + + +
CC = Crohn's colitis; UC = ulcerative colitis. b CAHTSDSgR indicates colonic segments from cecum (left) to rectum (right) (C = cecum; A = ascending; H = hepatic flexure; T = transverse; S = splenic flexure; D = descending; Sg = sigmoid; R = rectum); (+) denotes involvement and (-) absence of involvement of the particular segment. M = mild; Mod = moderate; MS = moderately severe; S = severe; None = no uptake. C Letters indicate study before (a) and after (b) therapy for acute colitis. d CDAI score could not be estimated (see text).
between the clinical assessment (CDAI score) and the indium scan grading (r = 0.81) is shown in Figure 2. In no instance did the scan fail to show radionuclide uptake when the colitis was clinically active. Two patients (cases 2 and 5) had indium scans during acute colitis and repeat scans when asymptomatic. In both cases, radionuclide uptake in the colon changed from universal uptake to no uptake whatsoever. Eleven of 17 scans (65%) were completed on patients who were taking some form of corticosteroid. The dosage of oral prednisone (average 45 mg) or intravenous methylprednisolone (averge 65 mg), although relatively high, did not appear to inhibit leukocyte migration to the inflamed colon. Although the leukocyte scan appeared to readily Table 2. Correlation of Indium Scan and Other Studies for Disease Location Agreement category Perfect Good Moderate Poor
No. of discrepant colon segments O None 1 2-3 3
No. of scans in categoryb 8 3 2 3
(50.5%) (18.5%) (12.5%) (18.5%)
°Number of colon segments in disagreement between scan and other studies (Barium enema x-rays; colonoscopy; surgical specimen). b Sixteen scans in 15 patients; patient 2 studied twice and patient 5a not included (See text and Table 1.)
identify colonic inflammation, it did not reliably differentiate ulcerative and Crohn's colitis. The diagnosis was usually apparent, however, after clinical assessment and sigmoidoscopic examination.
Discussion The indium 111 leukocyte scan has several unique characteristics that favor its use as a diagnostic modality in acute ulcerative and Crohn's colitis. The test is noninvasive, easily tolerated, simple to perform, and utilizes commonly available gammascanning equipment. The radionuclide hC!s emission peaks and a physical half-life that make it ideal for clinical use (17,24), and indium lll-labeled leukocytes show no uptake in the normal gastrointestinal tract (14,25). These qualities allow its use during active or recurrent colitis and in patients who are too fragile for invasive studies. All of the patients in this study had abnormal radionuclide uptake in the colon during episodes of active colitis. The scan findings correlated well with those obtained by established techniques. Where the correlation was poor, there was usually a lengthy delay between the white blood cell scan and the subsequent radiographic examination. Other explanations for discrepant findings include the tendency for labeled leukocytes shed into the bowel lumen to migrate with the fecal stream to areas where no disease is present, and recognition that the scan may
392
STEIN ET AL.
delineate areas of inflammation not apparent on barium enema. Direct observation of the mucosa by colonoscopy will often detect such areas. It should also be recognized that labeled leukocytes will not migrate to regions of inactive disease whereas these regions might be readily apparent with other diagnostic techniques. There is no established grading system for clinical disease activity that can be applied to both ulcerative and Crohn's colitis. The CDAI (21,22) accurately grades Crohn's disease, but has not been used in ulcerative colitis. The parameters that are measured to obtain the CDAI score, however, are also major indicators of disease activity in ulcerative colitis (sense of well-being, abdominal pain, number of watery or bloody stools per day, constitutional symptoms, and extracolonic manifestations of disease). Since we felt it important to have a common grading system for both diseases, the CD AI was chosen because it is an established system in Crohn's disease and because it correlated well with the physician's subjective assessment of activity in ulcerative colitis. In only 1 case did the CDAI misrepresent disease activity; that occurred in a mentally retarded quadriplegic patient who was unable to respond to the appropriate CDAI questions (case 7). The indium 111 leukocyte scan identified active colitis, whether Crohn's or ulcerative, and whether new or recurrent in onset. In addition, the use of corticosteroids did not appear to inhibit the migration of labeled cells to the diseased colon. There was a remarkably good correlation between the clinical assessment of disease activity and the degree of radionuclide uptake. This suggests that the indium scan may be useful in following the course of disease activity. As anticipated, the leukocyte scan did not distinguish between the two types of colitis. Patchy or right-sided uptake was evident in some cases of Crohn's disease, but no specific diagnosis could be made in most patients with universal colitis. Yet, the indium scan adds a diagnostic dimension to sigmoidoscopy, barium enema, and colonoscopy. If a diagnosis can be made through clinical assessment and sigmoidoscopy, then the scan offers an easy noninvasive method to define the extent and activity of disease. No significant complications were encountered during the use of this diagnostic technique. No colon preparation was necessary before scanning, and all of our patients tolerated the procedure easily. The severity of colitis was not intensified by the study and we observed no cases of megacolon, perforation, sepsis, or hemorrhage that were temporally related to radionuclide imaging. The total whole body radiation exposure with the
GASTROENTEROLOGY Vol. 84. No.2
600
•
x w o z
>-
~
:> ~
u <[ ....J
<[
U Z
....J U
o~------~------~------~------~
o
2
III
3
4
In - SCAN SCORE
Figure 2. Correlation of clinical estimation of disease activity (CDAI. see Methods) and indium scan interpretation of activity (1 = mild. 2 = moderate. 3 = moderately severe. and 4 = severe).
indium 111 leukocyte scan is approximately 0.31 rad/500 JLCi dose, with the highest concentrations seen in the liver (1.3 rads/500 JLCi) , spleen (8.5 rads/500 JLCi), and bone marrow (2.3 rads/500 JLCi) (22). A barium enema with fluoroscopy and spot films gives a slightly greater whole body dose, but less intense exposure to the liver, spleen, and bone marrow (26). A standard gallium 67 scan (5 mCi dose) gives comparable exposure to these organs. Although previous pr~liminary reports have shown positive indium scans in inflammatory bowel disease (27), our study extends those findings by identifying patients prospectively and comparing the scan results with an assessment of clinical status and with the findings on barium enema or colonoscopy. This diagnostic technique may be of value in defining the location of colitis early in the course of acute disease when barium contrast studies are contraindicated. Its eventual clinical application will depend on future experience with larger numbers of patients in a variety of clinical situations. Although its role in the management of inflammatory bowel disease remains to be established, the indium scan
February 1983
appears to have great potential in assessing the activity and extent of Crohn's and ulcerative coliti~
INDIUM 111 LEUKOCYTE SCAN IN COLITIS
15.
References 1. Edwards FC, Truelove C. The course and prognosis of ulcerative colitis. Gut 1963;4:299-315. 2. Wruble LD, Bronstein MW. Toxic dilatation of the colon following barium enema examination during the quiescent stage of chronic ulcerative colitis. Am J Dig Dis 1968;13:91823. 3. Binder SC, Patterson JF, Glotzer DJ. Toxic megacolon in ulcerative colitis. Gastroenterology 1974;66:909-15. 4. Jalan KN, Sircul W, Card WI, et al. An experience of ulcerative colitis. l. Toxic dilatation in 55 cases. Gastroenterology 1969;57:68-82. 5. Norland CC, Kirsner JB. Toxic dilatation of colon (toxic megacolon): etiology, treatment and prognosis in 42 patients. Medicine 1969;48:229-50. 6. Waye JD. The role of colonoscopy in the differential diagnosis of inflammatory bowel disease. Gastrointest Endosc 1977;23:150-4. 7. Myren J. Eie H, Serck-Hanssen A. The diagnosis of colitis by colonoscopy, with biopsy and x-ray examination. A blind comparative study. Scand J GastroenteroI1976;11:141-4. 8. Thakur ML, Coleman RE, Welch MJ. Indium-lll-labeled leukocytes for the localization of abscesses: preparation, analysis. tissue distribution and comparison with gallium-67 citrate in dogs. J Clin Lab Med 1977;89:217-28. 9. Thakur ML, Coleman RE, Mayhall CG, Welch MJ Jr. Preparation and evaluation of 111 In-labeled leukocytes as an abscess imaging agent in dogs. Radiology 1976;119:731-2. 10. Zakhireh B, Thakur ML, Malech HL, Cohen MS, Gottschalk A. Root RK. Indium-l11-labeled human polymorphonuclear leukocytes: viability, random migration, chemotaxis, bactericidal capacity, and ultrastructure. J Nucl Med 1979;20:741-7. 11. Thakur ML, Lavender JP, Arnot RN, Silvester DJ, Segal AW. Indium-lll-labeled autologous leukocytes in man. J Nucl Med 1977;18:1014-21. 12. Segal AW, Thakur ML, Arnot RN, Lavender JP. Indium-l11labeled leukocytes for localization of abscesses. Lancet 1976;ii:l056-8. 13. McDougall IR. Baumert JE, Lantieri RL. Evaluation of l11In leukocyte whole body scanning in 59 patients. Am J Roentegenol 1979;133:849-54. 14. Goodwin DA, Doherty PW, McDougall IR. Clinical use of indium-l11 labeled white cells. An analysis of 312 cases of
16. 17. 18.
19. 20.
21.
22. 23.
24.
25. 26.
27.
393
radiolabeled blood cellular elements. In: Gottschalk A, Thakur ML, eds. New York: Trivirum Publishers, 1981:131-45. Winzelberg GG, Castronovo FP, Callahan RJ, McKusick KA, Strauss HW. 111In oxine labeled red cells for detection of simulated lower gastrointestinal bleeding in an animal model. Radiology 1980;135:455-61. Weibel BJ, Forstrom L, McCullough J. Studies of the kinetics of indium-l11-labeled granulocytes. J Lab Clin Med 1979;94:246-55. Dutcher JP, Schiffer CA, Johnston GS. Rapid migration of l11indium-labeled granulocytes to sites of infections. N Engl J Med 1981;304:586-9. Ferrant A, Dehasque N, Leners N, Neunier H. Scintigraphy with In-l11-labeled red cells in intermittent gastrointestinal bleeding. J Nucl Med 1980;21:844-5. Thakur ML. Indium-lll: a new radioactive tracer for leukocytes. Exp Hematol 1977;5(suppl):145-50. Sparberg M, Fennessy J, Kirsner JB. Ulcerative proctitis and mild ulcerative colitis: a study of 220 patients. Medicine 1966;45:391. Best WR, Bectel JM, Singleton JW, Kern F Jr. Development of a Crohn's disease activity index. National Cooperative Crohn's Disease Study. Gastroenterology 1976;70:439-44. Best WR, Bectel JM, Singleton JW. Rederived values of the eight coefficients of the Crohn's disease activity index (CDAI). Gastroenterology 1979;77:843-6. Goodwin DA, Finston RA, Smith Sl. The distribution and dosimetry of In-111-labeled leukocytes and platelets in humans. Proceedings of the Third International Radiopharmaceutical Dosimetry Symposium. Oak Ridge, Tenn: Food and Drug Administration, 1980:88-101. Goodwin DA, Sundberg MW, Diamanti CI, Meares CF. l11In_ labeled radiopharmaceuticals and their clinical use. In: Subramanian GS, Rhodes BA, Cooper JF, and Sodd VJ, eds. Radiopharmaceuticals. New York: Society of Nuclear Medicine, 1975:80-101. Saverymuttu SH, Peters AM, Lavender JP, Hodgson HJ, Chadwick VS. Indium 111 labelled autologous leukocytes in inflammatory bowel disease (abstr). Gastroenterology 1981;80:1273. Shleien B, Tucker TT, Johnson DW. The mean active bone marrow dose to the adult population of the United States from diagnostic radiology. Washington, D.C.: U.S. Department of Health, Education, and Welfare, Bureau of Radiological Health, January 1977. Segal AW, Munro JM, Ensell J, Sarner M. Indium-lll tagged leukocytes in the diagnosis of inflammatory bowel disease. Lancet 1981 ;ii:230-2.