90 Gut 1994; 35:90-93 in Crohn's disease and indications for its assessment using SeHCAT Gut: first published as 10.1136/gut.35.1.90 on 1 January 1994. Downloaded from

H Nyhlin, M V Merrick, M A Eastwood

Abstract theless, some authors have reported a high Patients with Crohn's disease who suffer from prevalence.45 longstanding diarrhoea that does not respond Conventional methods of diagnosing diseases to conventional treatment pose a common of the terminal lack either sensitivity or clinical problem. Bile acid malabsorption is a specificity, or both. Radiological studies give no possible cause, although its prevalence and information about functional status, while the clinical importance is unclear. This paper Schilling test is influenced by factors not con- explores the clinical indications for referring nected with terminal ileal function, such as patients with Crohn's disease for bile acid bacterial colonisation and pernicious anaemia. assessment and the extent of bile acid mal- The cholyl-glycine-l-14C breath test measures absorption in this selected group of patients. terminal ileal function but is also influenced by The selenium labelled bile acid SeHCAT was bacterial contamination of the small bowel, the used to assess the effect of disease on the latter being the principal indication for perform- integrity of the enterohepatic circulation. ing this test. Direct measurement of faecal bile Altogether 76% ofthe patients referred for bile acids or of faecally excreted14C-labelled conju- acid assessment had longstanding diarrhoea gated bile acids are sensitive tests giving accurate that had not responded to conventional anti- measurements of the functional condition of the diarrhoeal treatment or an increase in steroid terminal ileum, but they are laborious, rely on therapy as their sole or predominant symptom. faecal collections, and are not easily performed in Ninety per cent of patients with bowel resec- patients with diarrhoea. The introduction of a tions, almost exclusively ileocaecal, had selenium labelled bile acid - SeHCAT67 - pro- abnormal SeHCAT retention (<5% at seven vided a simple and reliable method of studying days). Twenty eight per cent of patients with the enterohepatic circulation and the effect of Crohn's disease who had not undergone resec- diseases on its integrity. We report here our

tion 28% had a SeHCAT retention <5%, conclusions concerning its role in the clinical http://gut.bmj.com/ signifying bile acid malabsorption. Nineteen of management of patients with Crohn's disease, 22 patients given cholestyramine treatment after six years experience. subsequent to the SeHCAT test had a good symptomatic response. In conclusion, the pre- valence of bile acid malabsorption in this Patients and methods selected group with Crohn's disease is suffici- Between 1983 and 1989 53 patients with Crohn's

ently high to justify performing the SeHCAT disease from two hospitals in Edinburgh that on September 28, 2021 by guest. Protected copyright. test in order to separate the various differential have specialist gastrointestinal facilities were diagnoses. investigated for bile acid malabsorption using (Gut 1994; 35: 90-93) SeHCAT. This test was not routinely available before 1983. There was no agreed protocol and patients were investigated only to the extent Chronic or intermittent diarrhoea in adults may considered clinically appropriate by the clinician be an expression of various intestinal infections, concerned. SeHCAT retention was measured inflammatory bowel disease, malabsorption, or only after other investigations had failed to functional disorders of the gastrointestinal account adequately for their refractory diar- system. Aqueous concentration of bile acids in rhoea. Patients who had been referred for the colon greater than 3 mM, especially of SeHCAT retention measurement were identified chenodeoxycholic or deoxycholic acids, whether from the records of the Department of Nuclear due to disease, after resection of the terminal Medicine and their notes were reviewed in 1990 ileum, or idiopathic, may produce diarrhoea' as a by an independent physician (HN). Clinical consequence of a number of factors operat- details before and after investigation were avail- ing synergistically. These include increased able in all but two of the patients, allowing for Gastro-Intestinal Unit epithelial permeability resulting from a deter- assessment ofthe reasons for investigation, local- and Department of gent effect, stimulation of cyclic adenosine isation of disease, and the implications of the Nuclear Medicine, monophosphate (cAMP) in the colonic mucosa, SeHCAT results for treatment. Western General Hospital, Edinburgh nervous mechanisms, and/or increased colonic Twenty five of the patients had unoperated H Nyhlin motility. The effect is exacerbated in the Crohn's disease and their symptoms had failed to M V Merrick presence of pronounced steatorrhoea by certain respond adequately to conventional treatment. M A Eastwood metabolites of fatty acids produced by colonic The distribution of disease is shown in Table I. Correspondence to: Dr M A Eastwood, bacteria. The prevalence and clinical importance Twenty six had previously undergone bowel Gastro-Intestinal Unit, of bile acid malabsorption in Crohn's disease is resection for their Crohn's disease (22 ileocaecal, Western General Hospital, Edinburgh EH4 2XU unclear. It is usual after resection of the terminal three colonic, and one limited ileal resection. Accepted for publication ileum, but according to some reports occurs in Thirty one were female (mean age 39-8 years, 12 May 1993 only a minority of unoperated patients.2 3 Never- range 16-72 years) and 20 male (mean age 37-9 Bile acid malabsorption in Crohn's disease and indicationsfor its assessment usingSeHCA T 91

years, range 16-60 years). The mean duration of TABLE II Localisation in relation to SeHCAT retention disease was 9-6 years (range 1-31 years). Localisation After resection SeHCAT retention seven days after oral without

was measured using a shadow- Small radiological Gut: first published as 10.1136/gut.35.1.90 on 1 January 1994. Downloaded from administration intestinal Small signs of shield, whole body gamma counter, as pre- SeHCAT +colonic intestinal Colonic recurrence viously described.8 Patients attended in the retention (n=9) (n= 15) (n= 14) (n = 13) morning, having fasted overnight. After <5% 5 12 1 12 measurement of their background radioactivity 5-10% 1 - 2 1 levels (principally to ensure that they had not >10% 3 3 11 received other radioactive tracers), 40 kBq of 58CoB12 was administered orally with a drink of water. Thirty minutes later the patient's radio- TABLE III SeHCAT results in relation to reasonfor activity level was counted again to obtain the investigation (all patients) 100% count rate value and also an estimate ofthe Diarrhoea+ scattered counts from the cobalt in the selenium SeHCAT Diarrhoea extension of Diarrhoea+ window. A total of 40 kBq of 75SeHCAT was retention only disease miscellaneous then administered orally, and after a further wait <5% 25 1 4 of half an hour, a second measurement was 5-10% 3 - 1 made. The contribution of the scatter from 58Co >10% 11 6 4 in the 75Se window was calculated and sub- tracted. The patients returned one week later, measurement was made of the when a further TABLE IV Outcome ofSeHCAT in Crohn's disease patients radioactivity in both selenium and cobalt with and without bowel resection windows. The retention ofeach was expressed as a percentage of the initial administered activity. Bowel resection No resection The in vivo stability of both SeHCAT9 and SeHCAT retention (n=26) (n=25) are well established. Phanton measure- <5% 23 7 58CoB12 5-10% 3 1 ments in our department indicate that the tech- > 10% 3 17 nique has a precision ofbetter than +005% and a sensitivity of better than 1% of the administered activity. TABLE V Cholestyramine, effectiveness on symptoms in relation to SeHCAT results Results SeHCAT retention Effective Not effective The most common indication for investigation http://gut.bmj.com/ <5% 18 1 (Table I) was the presence of diarrhoea that had 5-10% - 2 failed to respond to conventional treatment and >10% 1 - for which no adequate diagnosis had been made. Thirty one ofthe 51 patients (76%) had diarrhoea as their predominant or only symptom and in most cases it was of lengthy duration (greater SeHCAT retention of less than 5%; he also had than three months) and had not responded to ileal strictures on barium follow through. Nine on September 28, 2021 by guest. Protected copyright. conventional antidiarrhoeal treatment or to an patients who had severe symptoms in addition to increase in steroid therapy. SeHCAT retention their diarrhoea were referred for investigation of was less than 5% at seven days in 25 of these suspected bile acid malabsorption, including two patients (64%) and between 5% and 10% in three whose predominating symptom was flatulence. others (Tables II and III). In a further seven Both the latter two patients had SeHCAT reten- patients the diarrhoea had been controlled to tion of less than 5% at seven days and responded some extent by treatment, but there was a symptomatically to cholestyramine. suspicion that involvement ofthe terminal ileum Table IV summarises the SeHCAT results in was causing bile acid malabsorption and diar- relation to bowel resection. Twenty three of the rhoea. Four of these patients had a normal 26 patients (90%) with bowel resection had a barium follow through, in one this was not SeHCAT retention ofless than 5% at seven days. performed, and the two others had ileal stric- Twenty two of the 23 had undergone ileocaecal tures. Only one of the seven had a seven day resection and one had had a limited ileal resec- tion. The remaining three patients, with reten- tions of between 5% and 10%, had undergone limited colonic resections. In the 25 patients with TABLE I Localisation in relation to the indicationfor investigation in 51 patients with Crohn's Crohn's disease who had not had bowel resec- disease tion, seven (28%) had a SeHCAT retention of less than 5% at seven days, in one retention was Localisation After resection without between 5% and 10%, and 17 (68%) had no Small radiological evidence of bile acid malabsorption. Twenty two intestinal Small signs of +colonic intestinal Colonic recurrence patients were given cholestyramine subsequent ReasonsforSeHCAT (n=9) (n= 15) (n= 14) (n= 13) to the SeHCAT test, with a good symtomatic Diarrhoea only (n= 39) 8 (4) 1 1 9 11 response in 19 (Table V). B12 absorption is Diarrhoea and clinical suspicion ofextension of 1 1 5 - measured routinely in our protocol as an disease (n = 7) Diarrhoea plus other factors (incontinence (2), 1 3 3 2 indicator of bacterial overgrowth. All the weight loss (2), pre-op (1), flatulence (2), routine patients investigated had values within the check up (1), perianal fistula (1) normal range (retention >23%). 92 Nyhlin, Merrick, Eastwood

Discussion competition for uptake of vitamin B12, reflected SeHCAT had been available for 10 years, but has in a low 58CoBl2 retention. There is no single test been slow to obtain wide acceptance. Many of that will detect all the organisms which may

the initial studies concentrated on its use in colonise the small bowel. In this study, no Gut: first published as 10.1136/gut.35.1.90 on 1 January 1994. Downloaded from patients with small bowel resections, in whom bacterial overgrowth could be shown. This may the a priori probability ofbile acid malabsorption seem surprising at first glance as bacterial over- is very high. In these patients it rarely added growth is common in patients with Crohn's useful clinical information. Our experience in disease localised to the ileocaecal region. It may Edinburgh, where the test was developed and be explained synergistically by a number of first employed, remains the largest yet reported, factors. The most important is almost certainly with over 1000 patients investigated for a variety selection. As is usual, clinical investigations were ofconditions. The accuracy and sensitivity ofthe performed hierarchically, not inclusively, those SeHCAT test for detecting bile acid malabsorp- for the commoner conditions taking precedence. tion is now well established.'° A particular Tests of colonisation thus preceded SeHCAT advantage of this test over 14C glycocholate and retention determination, which was not per- 14C taurocholate is its ability to differentiate formed if a diagnosis had been obtained before between malabsorption and colonisation, as reaching this level on the diagnostic 'tree'. Thus, SeHCAT is largely resistant to bacterial decon- patients with appreciable colonisation are likely jugation and dehydroxylation. It is thus a pure to have been excluded. Secondly, the extent of marker of active transport of bile acids, which small bowel involvement sufficient to cause bile under most circumstances is the predominant acid malabsorption is commonly insufficient to mode ofabsorption and is limited to the terminal give detectable B12 malabsorption.'2 Thirdly, one metre of ileum, as opposed to passive there are important biological differences in diffusion, which occurs throughout the small behaviour and handling between SeHCAT and and large bowel to an extent determined by the 58CoB12. Whole body retention of58CoBl2 repre- polarity of the particular bile acid or conjugate. sents largely the amount absorbed from a single Free bile acids are not present in the small passage across the ileum, and once absorbed it is intestine unless there is substantial colonisation not excreted in appreciable amounts. In con- with certain anaerobic bacteria. There is little trast, after ileal absorption that has an efficiency passive diffusion oof the conjugates normally of 95%-98%, SeHCAT is carried to the , found. In particular the role of SeHCAT in the extracted, and secreted in the bile from which it of the diarrhoeal form of re-enters the enterohepatic cycle. With this has recently been compound, therefore, the ileum is tested several clarified." 12 The presence of bile acid mal- times a day, while SeHCAT itself is effectively

absorption in patients with a diseased but intact resistant to the microbiological deconjugation. http://gut.bmj.com/ ileocaecal region could well be transient, chang- This has the effect of amplifying a moderate ing with disease activity in the early phase of the absorptive defect which may remain undetected disease, although this has not been documented. using the single-passage absorption of 58CoB12. With time, however, and repeated relapses, This study was limited to a selected subgroup fibrosis is likely to develop, resulting in a perma- ofpatients with a high a priori probability of bile nent impairment ofbile acid absorption. acid malabsorption. They had all failed to

The purpose of the present study was, firstly, respond to conventional treatment. Significant on September 28, 2021 by guest. Protected copyright. to explore the (de facto) clinical indications for bile acid malabsorption was mainly found in referring patients with Crohn's disease for bile operated patients and only in a relatively small acid assessment in a centre where this test was number of unoperated patients with Crohn's familiar and secondly, to assess the extent ofbile disease. From the present series it is clear that acid malabsorption among this selected group of SeHCAT, used selectively in patients with patients. Some previous studies'3"- selected Crohn's disease to investigate longstanding diar- small series of patients with Crohn's disease on rhoea that has failed to respond to conventional the presumption that these patients were likely treatment, commonly identifies a treatable cause to have bile acid malabsorption because of the (39 of 51, 76%). Nineteen ofthe 25 patients with common involvement of the absorptive sites of bile acid malabsorption had undergone bowel bile acids in the terminal ileum. In view of the resection, leaving six patients who had not been patchy distribution of activity in Crohn's operated on for their Crohn's disease, but who disease, however, it is not surprising that a poor nevertheless had bile acid induced diarrhoea. correlation was found between the severity of Thus, although the cause of the diarrhoea in the bile acid malabsorption and radiological and 25 unoperated patients was presumed clinically other criteria ofdisease activity. to be bile acid malabsorption, in only seven The three basic mechanisms of B12 deficiency (23%) was malabsorption significant, indicating are well known: deficiency of intrinsic factor that this is not a common cause of longstanding associated with pernicious anaemia, utilisation of diarrhoea in these patients. It is, however, causal B12 by bacteria in a colonised , and in a significant minority who fail to respond to loss of the specific transport sites due to disease usual first line treatment. It is clear from these or resection oftht erminal ileum. The use of results that bile acid malabsorption occurs dual isotopes - free B12 labelled with 58Co and mainly in patients who have had an ileocaecal IF-Bl2 with 57Co - allows simultaneous determi- resection and is infrequently the cause of notice- nation of the level of defect, whether gastric or able symptoms in unoperated patients. As absorptive. In this study only 58Co B12 was patients were not routinely investigated after administered as interest was focused on ileal resection, even the operated group is a selected absorption. In bacterial overgrowth there is cohort in whom there was uncertainty whether Bileacid malabsorption in Crohn's disease and indicationsfor its assessment usingSeHCA T 93

the extent of resection was adequate to account treatment, SeHCAT commonly identifies a for the symptoms. This underlines the import- treatable cause (39 of 51, 76%). ance ofclose cooperation between physicians and

surgeons in the follow up ofthese patients. 1 Hofmann AF. The enterohepatic circulation of bile acids in Gut: first published as 10.1136/gut.35.1.90 on 1 January 1994. Downloaded from health and disease. In: Sleisenger MH, Fordtran JS, eds. When bile acid malabsorption has been diag- . London: WB Saunders, 1989. nosed, a good therapeutic response is usually 2 Otte JJ, Andersen JR. The clinical value of faecal bile acid determination in patients with chronic diarrhoea of obtained to cholestyramine, and this is some- unknown origin. ScandJ Gastroenterol 1986; 21: 585-8. times used as evidence of the disorder. By 3 Scarpello JHB, Sladen GE. '4C glycocholate test in Crohn's disease - its value in assessment and treatment. Gut 1977; 18: slowing large bowel transit, however, cholestyra- 736. mine may act non-specifically. A therapeutic 4 Farivar S, Fromm H, Schindler D, McJunkin B, Schmidt KW. Tests of bile acid and vitamin B12 metabolism in ileal response in isolation is weak evidence only ofbile Crohn's disease. AmJ Clin Pathol 1980; 73: 69-74. acid malabsorption. The single patient who 5 Van Blankenstein M, Hoyset T, Horchner P, Frenkel M, Wilson JHP. Faecal bile acid radioactivity a sensitive and apparently responded to cholestyramine despite relatively simple test for ileal dysfunction. Netherlands a normal SeHCAT may well have had a spon- Joumal ofMedicine 1977; 20: 248-52. 6 Merrick MV, Eastwood MA, Anderson JR, Ross HMcL. taneous cyclical change in clinical severity, while Enterohepatic circulation in man of a gamma-emitting bile- the failure of response to cholestyramine in one acid conjugate, 23-selena-25-homotaurocholate. J NuclMed 1982; 23: 126-30. patient is thought to be due to non-compliance. 7 Nyhlin H, Merrick MV, Eastwood MA, Brydon WG. Evalua- In other patients the effect of cholestyramine on tion ofileal function using 23-selena-25-homotaurocholate, a gamma-labelled conjugated bile acid. Initial clinical assess- symptoms is impressive. It may well be that the ment. 1983; 84: 63-8. case for a compliant medical regimen can be put 8 Merrick MV, Eastwood MA, Ford MJ. Is bile acid malabsorp- tion underdiagnosed? An evaluation ofaccuracy ofdiagnosis forward more strongly and more confidently by measurement of SeHCAT retention. BMJ 1985; 290: when a definite diagnosis of bile acid malabsorp- 665-8. 9 Monks R, Boyd GS. Biologic stability of tauro-23-[75Se] tion has been made, and a clinically attractive selena-25-homocholic acid.J NuclMed 1988; 25: 1411-8. method such as the SeHCAT test may serve this 10 Merrick MV. Bile acid malabsorption. Clinical presentation and diagnosis. DigDis 1988; 6: 159-69. purpose. 11 Eastwood MA, Merrick MV. Bile acids in irritable bowel syndrome. In: Northfield TS, Jazrawi R. Bile acids in health and disease. Dordrecht: Kulver Academic, 1990: 267-73. 12 Williams AJK, Merrick MV, Eastwood MA. Idiopathic bile CONCLUSIONS acid malabsorption - a review if clinical presentation, diagnosis and response to treatment. Gut 1991; 32: 1004-6. Even though the study was limited to a selected 13 Ferraris R, Jazrawi R Bridges C, Northfield TS. Use of a subgroup of patients considered clinically to labelled bile acid (7 SeHCAT) as a test of ileal function. Methods of improving accuracy. Gastroenterology 1986; 90: have a high a priori probability of bile acid 1129-36. malabsorption, significant malabsorption was 14 Ordholm M, Pedersen JO, Arnfred T, Rodbro P, Thaysen EM. Evaluation of the applicability of the SeHCAT test in found not only in a minority of unoperated the investigation of patients with diarrhoea. Scand J patients with Crohn's disease who failed to Gastroenterol 1988; 23: 113-7.

15 Fagan EA, Chadwick VS, Baird IMcL. SeHCAT absorption: http://gut.bmj.com/ respond to conventional therapy, but in a larger a simple test ofileal dysfunction. 1983; 26: 159-65. group of operated patients. In patients with 16 Nyhlin H, Brydon WG, Danielsson A, Westerman S. Clinical application of a (75Se)-labelled bile acid for the investigation Crohn's disease who have longstanding diar- of terminal ileal function. Hepatogastroenterology 1984; 31: rhoea that has failed to respond to conventional 187-93. on September 28, 2021 by guest. Protected copyright.