Evaluation o f SeHCAT test in detern1ining ileal involvement and dysfunction in Crohn's disease

Evaluation of SeHCAT test in determining ileal involvement and dysfunction in Crohn's disease. Can J Gastroenterol 1993; 7 ( 8) :597-601. In view of the imperfect methodology that exists for the detection of terminal ileal disease, the aurhors examined the ability of the SeHCA T test to detect ileal involvement and dysfunction in patients with inflammatory bowel disease. An attempt was also made to correlate the 23-selena-25-homotaurocholate retention (SeHCAT) test with disease activity, extent of ilea! disease, presence of diarrhea and response to cholestyraminc. Forty-three patients were studied, including 12 controls - 22 with Crohn's disease limited to the small bowel and nme with ulcerative or granulomatous colitis. The mean Se HCA T retention values were 25, 4 and 4 7%, respectively, all significantly different from each other (P<0.05). This gave a sensitivity of 91 %, a specificity of 81 % and an accuracy of 86%. No correlation could be established between the SeHCAT retention value and disease activity, extent of ileal disease, presence of diarrhea or response to cholestyramine.

C ROHN'S DISEASE IS A CHRONIC inflammatory disorder potentially involving any level of the gastrointestinal tract but with a predilection co involve the terminal ileum. The methods that presently exist for detecting disease of the terminal ileum are imperfect. The Schilling's test is neither sensitive for limited ileal disease nor specific (1 ). The Cl4-conjugated bile acid excretion with fecal collection is more specific and sensitive than the bile acid breath test (2)(3)(4)(5)(6); however, it is difficult as a routine clinical test. Barium examination underestimates the presence of terminal ileal disease and is not a rest of ilea! function (7-9). Endoscopic examination and biopsy of the terminal ileum are technically demanding and, again, are not tests of ilea! function.
An available method is the selenium-labelled synthetic conjugated trihydroxy bile acid called 75-SeHCAT (75 se-23-selena-25 homotaurocholic acid) (10,11). The absorption and excretion of SeHCA T is similar co cholic acid being absorbed by the terminal ileum and excreted solely by the biliary system. By virtue of the gamma ray emissions, it can be measured by whole body counting and by uncollimated gamma camera measurements (12). It Les valeurs de retention moyenne du SeHCAT ont ete entre 25,4 % et 47 % respectivement, touces nettement differentes les unes des autres (P<0,05) . Cela a conduit a un degre de sensibilite de 91 %, de specificite de 81 % et de precision de 86 %. Aucune correlation n'a pu etre etablie entre la valeur de retention du SeHCA T et l'activite de la maladie, l'etendue de l'atteinte ileale, la presence de diarrhee ou la reponse a la cholestyramine.
has been proposed that measurement of SeHCA T retention is a simple and acceptable test for investigating ilea[ function, especially in inflammatory bowel disease (13)(14)(15)(16)(17)(18)(19)(20). The purpose of this study was to evaluate the ability of the SeHCA T test to determine ilea[ involvement and dysfunction in patients with inflammatory bowel disease. An attempt was also make to correlate the SeHCA T retention with the extent of terminal ileal disease, the Crohn's disease activity index (CDAI), the presence of diarrhea and its response to cholestyraminc, a bile acid resin used in the treatment of bile salt catharsis.

SUBJECTS AND METHODS
The study was approved by the University of Saskatchewan Advisory Committee of Ethics in Human Experimentation. All subjects gave informed signed consent. Study population: Forty-three subjects were studied (26 males an<l 17 females), with age ranging from 25 to 69 years (mean37). Of these 43, 12 were healthy controls (10 males and two females) with no symptoms to suggest gastrointestinal disease, no history of biliary or intestinal surgery and normal bowel habits. A normal bowel habit was defined as passage of formed stool fewer than three times a Jay with no recent change, and with the person considering their habit as normal. The ages of these controls ranged from 24 to 42 years (mean 32).
Twenty-four of the subjects, aged from 22 to 56 years ( mean 3 1 ), were diagnosed as having Crohn's disease (confirmed by positive biopsy and/or typical radiological appearance). Of these, two had disease limited to the colon as verified by radiology and colonoscopy with ileoscopy, eight had active terminal ileitis and 14 had prior terminal ileal resection (10 of these with recent recurrence proximal to the ileocolic anastomosis). Presence and extent of terminal ileal involvement or removal was determined by radiological, surgical and/or pathologic means.
Seven patients, aged from 25 to 69 years (mean 37), had left-sided or panulcerative colitis proven by biopsy, and appropriate endoscopic and radiological studies. Of these seven, four had clinically active disease. Procedures: The isotope procedure included the oral administration of 0.37 MBq of 75 SeHCAT in capsule form after a 3 h fast. The initial count was measured 3 h after the intake of the isotope with the patient supine and then prone beneath an uncollimated gamma camera at a distance of 42 cm from the table top, centred to the patient's umbilicus. The activity of the 75 SeHCA T was measured using a 20% window around the 266 KeV photon peak. All counts were for 300 s. The geometric mean of the anterior and posterior counts was then calculated. The mean of the two background counts was subtracted from the geometric mean to give the final patient count. The second measurement was performed on day 7 in a similar fashion with no intervening dietary or activity changes. The measuremenr was corrected for decay and a percentage of the retained activity calculated with values equal to or greater than 19% was considered normal.
In those patients with known Crohn's disease, the CDAI was calculated for the week the test was being performed. All patients with clinical diarrhea (defined as unformed stool more than three times per day) were offered cholestyramine following the Sel !CAT study at an initial close of 4 g/day, increased to 8 g/day after one week if no response. The clinical response was compared with the SeHCAT value, with clinical response being defined as a stool becoming formed with a frequency of less than three times per day. A correlation was also sought between the SeHCA T test and the amount of the terminal ileum resected or involved with Crohn's disease.
Statistical analysis was carried out by using the nonparametric Mann-Whitney test of analysis.

RESULTS
The mean± SE SeHCA T retention for the control group of 12 normal individuals was 25.0±3.2%. In this group, three individuals had abnormally low SeHCA T retention values (14, 11 and 6%), representing apparent false positives. One individual with a false positive result was verified by repeat SeHCA T study (14 and 12%). In all three, the Schilling's test and radiological study of the terminal ileum were normal. All three continue to remain well without diarrhea or other intestinal symptoms after one year follow-up.
In patients with Crohn's disease, the overall mean SeHCA T retention value of those with terminal ilea[ disease and/or resection was 5.0±2.0%, significantly different from the control group (P<0.01 ). T his group included two false negatives with SeHCAT values of 20 and 29%; one patient had 11 cm of terminal ileum resected with 10 cm of recurrent proximal disease and the other had 25 cm of terminal ilea[ involvement. Neither had evidence of colonic involvement.
The mean SeHCA T value in patients with ulcerative o r granulomatous colitis was 4 7.5±6. 7% which was significantly h igher than the mean for the terminal ilea! disease group (P<0.01 ). It was also significantly higher than the mean for the control group (P<0.05) (Figure 1). In this group of col itis patients, there was o ne false positive, a patient labelled as left-sided ulcerative colitis whose SeHCAT value was 15% and, after a one-year follow-up, continues to demonstrate a radiologically normal terminal ileum and Schilling's test. The two highest values in this group (79 and 64%) were individuals with active panulcerative colitis.  Assuming the distal ileum to be 100 cm, a relationship was sought between the percentage retention of SeHCA T and the residual distal ileum. H owever, using correlation analysis, no correlation was found overall (Figure 2). Of the 11 patients with retained ileocecal valve, the mean SeHCA T retention was 8.7±3.7% with mean residual ilea! length of 68.0±7.5 cm (correlation coefficient =0.457, P=0.163 ). Five of these 11 patients had diarrhea. O f the 11 patients without ileocecal valve, the mean SeH CA T retention value was 1.1±0.8%, with a mean residual ilea! length of 48.0±8. l cm (correlation coefficient -0. 134, P=0.742). Nine of 11 patients had d iarrhea.
In patients with terminal ilea! dis-ease and/or resection , 20 of 22 patients also had a low SeHCA T retention value; of this group, 13 (65%) had d iarrhea while seven (35% ) did not. In the group with diarrhea, only seven (54%) respo nded to the cho lestyramine therapy. Thus, a low SeHCAT retention value (less than 19% ) in patien ts with terminal ilea! C rohn's d isease did not reliably predict the presence of diarrhea or the clinical response of those with diarrhea to cho lestyramine. There was no correlation established between the C DAl and the SeHCAT retention values (correlation coefficient -0.25). The mean of those with active Crohn's disease (CDAI more than 150) was 4.0±1.2% while the mean of those with inactive disease (CDAI less than 150) was 3.0±1.8% (Figure 3 ).
The mean length of terminal ilea! involvement/ resection in those with diarrhea was 45.8±4. l cm (range 25 to 7 5) which was significantly different

DISCUSSION
The SeHCA T test is a test of bile salt absorption and has been proposed as a test of ilea! function. In this study, it appears that in patients with C rohn's disease, a low SeH CA T retention value st rongly correlates with ileal involvement. H owever, the sensitivity of 91 % and specificity of 81 % for the SeHCA T test is this study are not as high as previous studies quoting as high as 94% sensitivity and 100% specificity. The reasons fo r this are unclear. In our study, a value of greater than 19% fo r SeHCAT retention at d ay 7 was con-sidered normal. Values between 12 and 19% h ave previously been defined as equivocal. If, for this study, the val ues of 12% o r lower were defined as abnormal, this would have the effect of reducing the sen s1ttv1ty to 82%, increasing the specific ity to 90% but not chan ging the accuracy of 85%. Recent observations claim improved sen-smv1ty with three-day retention readings or analysis of activity versus time monoexponential curves calculated by the least squares method. Recently shown is that d ata uncorrected for colonic retention overestimate SeHCA T absorption to a variable degree with a resulting reduction in accuracy of the test. This would likel y be of importance in those witho ut diarrhea.
Indeed, the true sensitivity and specificity of the SeHCA T test is difficult to determine as we do not have a gold standard for diagnosing ileal dysfunction against which to compare. H owever, the sensitivity and spec ific ity va lues in this study of patients with morpho logical ev i<lence of Crohn 's disease limited to the ileum are equa l to o r better than the reported values for other ilea! function tests, including the Schilling's test, bile acid breath test and the C1 4-conjugated bile acid stool collection. In view of this and of the simplicity of the test, the SeH CA T test could be considered a n appropri ate investigation for diagnosing ilea! involvement in Crohn 's disease, espec-ially in distinguishing those with isolated large bowel invo lvement from those with isolated ilea! involvement. The highest retention values were achieved by patients with active pancolitis. Although up to 30% of the daily bile sa lt load can be absorbed by the healthy colon, the handl ing of bile salts by the inflamed colon is unknown. In this study, given that the SeH CAT retention values are significantly e levated in active colitis versus control values, this implies an enhanced absorption of bile salts by the inflamed colon which could potentially lead to a false negative SeHCA T retention value in patients with both ilea! and extensive colonic involvement.
As would be expected, the SeHCAT values do not correlate with disease activity as measured by the CDAl, but correlates with the presence of ilea! involvement, whether active, inactive or surgically removed. No correlation could be established between the SeHCA T value and the amount of tenninal ileal remaining, which may represent a type lI error as prior studies have suggested such a relationship. As the majority of SeHCA T retention values for terminal ilea! disease/resection are vety low, it is possible that earlier readings on day 3 or calculation of the area under the retention curve may be more accurate in quantitating the rate of bile salt loss and, thus, more accurately correlate with the extent of ilea! disease/resection.
In agreement with prio r studies is Furthermore, a low SeHCA T retention value in those with diarrhea a nd with Crohn's disease limited to the terminal ileum is not clinically useful in predicting the clinical response to cholestyramine therapy. Clearly the SeHCAT test is able to de mo nstrate bile salt loss but the test is unable to establish a causal relationship between bile salt malabsorption and diarrhea. This is understandable given the multiple mechanisms in Crohn's disease besides bile salt catharsis that may contribute to diarrhea. Thus it appears simple r to give an empiric trial of cho lestyra mine than to investigate the symptomatic patient with a SeHCA T test to determine whether bile salt catharsis is of importance clinically.