sign in sign up
<<
Home , Chronic pancreatitis

1076 Gut, 1990,31, 1076-1079 Pancreatic function in Crohn's disease Gut: first published as 10.1136/gut.31.9.1076 on 1 September 1990. Downloaded from

J Hegnh0j, C P Hansen, T Rannem, Heidi S0birk, Lykke Bjerglund Andersen, J R Andersen

Abstract related to the duration of disease,4 but otherwise We investigated exocrine pancreatic function the groups of patients have been too small to in a population of patients with Crohn's draw any conclusion concerning possible disease in order to correlate the pancreatic relations between clinical parameters and pan- function with clinical and laboratory variables. creatic function. A total of 143 patients affected by Crohn's The present study was performed to investi- disease and 115 control subjects were studied. gate exocrine pancreatic function in a population AU had a Lundh meal test. As a group patients of patients with Crohn's disease and to correlate with Crohn's disease had significantly pancreatic function with clinical and laboratory decreased activity of both (p<002) variables. and lipase (p<0-001) in duodenal aspirates. In patients with Crohn's disease enzyme activi- ties were not correlated to duration of disease Patients and methods or to extent or localisation of previous bowel During a four year period every patient admitted resection. The lowest enzyme values were to the department with known or suspected found in patients with the most extensive Crohn's disease had a Lundh meal test.5 Crohn's bowel involvement, and they were significantly disease was diagnosed in 143 patients (59 males, lower (p<005) than in patients with disease 84 females, mean age 37, range 15-80 years). confined to the terminal . The differ- Diagnostic criteria were based on radiological, ences between enzyme values in other sub- endoscopic, and histological findings and clinical groups ofpatients were not significant. For the history. Eighteen patients were treated with patient group as a whole no correlation was prednisolone, 29 with salazosulfapyridine, and found between disease activity and enzyme 10 with both drugs. Two patients had a jejuno- values, but for the most uniform group of stomy; none had an ileostomy or colostomy; and patients, those with terminal , pancreatic none had a history ofprevious function was significantly lower (p<005) in or alcohol abuse. The mean duration of disease patients with moderate and severe disease was 76 months (range 1-684 months). Table I compared with patients with mild disease. gives the site of disease. http://gut.bmj.com/ Thus atleast two factors seem to be responsible As control subjects, 115 patients (65 females, for impaired pancreatic function in Crohn's 50 males, mean age 46 years, range 13-77 years) disease: firstly disease activity and secondly were investigated during the same period for localisation or extent ofdisease. other causes or preoperatively before ileal bypass Department of operation for obesity. None had pancreatic or P, Division of biliary disease, presently or previously, or pre- , on September 27, 2021 by guest. Protected copyright. Rigshospitalet, Decreased function of the exocrine has vious gastrointestinal except for simple University of been reported in 20-30% ofpatients with various appendectomy. None had mellitus or a Copenhagen, Denmark intestinal disorders.'2 In 1953 Dreiling3 found history of abuse. Table II lists the C P Hansen impaired pancreatic function in five of 21 diagnoses. T Rannem patients with regional or diffuse In patients with Crohn's disease activity was Heidi S0birk jejunoileitis. Others have reported impaired Lykke Bjerglund divided into three grades (mild, moderate, Andersen exocrine pancreatic function in about 50% of severe) on the basis ofclinical criteria used in the J R Andersen patients with Crohn's disease,24 though the Harvey Bradshaw index6: mild 0-2, moderate Correspondence to: numbers were small (four and 17 patients). 3-8, severe >8. Dr Jens Hegnh0j, Department It has been suggested that decreased pan- of Medicine P, Division of All patients and control subjects had a Lundh Gastroenterology, creatic function in patients with Crohn's disease test.5 After an overnight fast a duodenal tube was Rigshospitalet, DK-2100 is a non-specific effect of protein deficiency placed with the tip in the horizontal part of duo- control. A standard- AcceptedACcptedhafofor publication secondary to . In a small series of denum under fluoroscopic 6 November 1989 patients the decrease in pancreatic function was ised meal was given. The meal (Nutrizon standard, Nutricia Nordica A/S Denmark) con- sisted of 12 g protein, 4 g fat, and 36 g carbo- TABLE I Lundh test results in 143 patients with Crohn's disease grouped according to hydrate mixed with water to a total volume of localisation ofdisease. Median values (ranges) 300 ml. Collection of duodenal aspirate was per- formed by simple siphonage (0-60 minutes). Localisation of Amylase - Crohn's disease No. (KU/I) (KU/I) The samples were put on ice and stored at 200C until analysed for amylase and lipase. Volume Terminal ileum 43 740(184-1732)* 199(33-933) Colon (including ) 15 751 (49-1412)* 175 (26-727) and pH were measured. The median volume Ileum 11 682(126-2095) 205(51-408) in patients was 66 ml (range 20-242 ml) and Terminal ileum+colon 59 600 (5-1542)*t 166 (3-538)* Extensive 15 496 (65-1584)*t 123 (26-447)*t in control subjects 54 ml (range 20-230 ml). (ileum+ 4, colon+ileum+ The median pH in patients was 6-22 (range jejunum 8, extensive ileum+colon 3) 5 50-686) and in control subjects 6-23 (range *Activity lower than in control subjects (p<005). tActivity lower than in patients with lesions of the 5 50-7 44). terminal ileum (p<005). Lipase was assayed by the method described Pancreaticfunction in Crohn's disease 1077

TABLE II Diagnoses in 9929n I_I _ I I I I control subjects without pancreatic disease (n= 115) 0 Gut: first published as 10.1136/gut.31.9.1076 on 1 September 1990. Downloaded from 0 Chronic diarrhoea 30 67 0~~~~~~~~~~~~~~~~~~ Diverticular disease 3 Adiposity 5 c0 OK3w ;O Miscellaneous 10 Go 0o0~~~~~~~~~~~~0G0 1500 0 0 - 0 o -= =)0 tO m ~ ~~o~ 00 0 CU 030 0w00CKX1 n0. O0OO tD 0 0 fll____ o 0 0 am Go 0 750i CMo o0 00X 0o 0 00 a_ Go ao.0 00 0 o---- o o co o co -4-ae o co C-o o * a *a o0 0D 0 0 0 o 00 ow 0 o 00 mw 0o0o06 -a 0 o 0 000 0 0 0 e 0 0 0 . 0 0 0I 0 A J- I II Il I HI I I III I II III 111 1I * * Terminal Colon Colon and Ileum Extensive controls terminal ileus

Site and activity of disease Duodenal lipase activity (KU/I) in 115 control subjects and 143 patients with Crohn's disease grouped according to site and activity ofdisease. Median values (ranges). *less than mild activity (p

by Erlandson and Borgstrom.7 Amylase was ileum and colon these values were significantly determined by the method described by higher (p<005). Otherwise no significant Dahlqvist with the modifications described by differences were noted between patients with Thaysen et al.89 different sites of Crohn's disease. For the patients as a whole disease activity did http://gut.bmj.com/ not correlate with enzyme activity, but in the STATISTICS group of patients with isolated disease of the Correlation analyses and the Mann-Whitney U terminal ileum the patients with moderate or test for unpaired data were used for statistical severe disease activity had significantly lower analysis. enzyme activity than those with mild or severe disease (Table III, IV; Fig). Enzyme activities in patients receiving prednisolone or salazosulfa- Results pyridine, or both, were not significantly differ- on September 27, 2021 by guest. Protected copyright. As a group patients with Crohn's disease had ent, and furthermore were not different from significantly decreased activities of both amylase activities in patients who received no medical and lipase in duodenal aspirates. The median treatment. Given the different sites of disease, amylase activity for patients was 170 KU/l (range enzyme activities were higher (p<005) in 3-933 KU/1) and for control subjects 199 KU/l patients with disease in the terminal ileum within (47-629 KU/1) (p<0 02). The median lipase every degree of activity. activity for patients was 702 KU/l (5-2095 KU/1) and for control subjects 880 KU/l (177-1908 KU/1) (p<0-001). In both groups there was a Discussion significant correlation between amylase and In the present study, which comprises the largest lipase activities (p<0-001). published number of patients with Crohn's Neither patients with Crohn's disease nor disease who had a test of exocrine pancreatic control subjects showed a significant correlation function performed, a Lundh test showed between age and pancreatic function. In the significantly decreased exocrine pancreatic patients amylase and lipase activities were not function in a group of 143 patients with Crohn's correlated with the duration of disease or with disease compared with a control group. This the extent or localisation of previous bowel agrees with the trend in previous studies which resection. included only a few patients. Nutritional status judged by serum albumin In a previous study the impairment of pan- concentrations did not correlate with the result creatic function was related to the duration of of the Lundh test. disease in a group of 17 patients with Crohn's Enzyme values in subgroups of patients disease.4 We could not reproduce this finding. grouped according to localisation of disease are In contrast to the previously published studies shown in Table I. The highest values were found on this subject, the large number of patients in in patients with isolated disease of the terminal our study allowed us to relate the enzyme ileum, and compared with patients with-exten- activities to the localisation of Crohn's disease. sive disease and disease confined to the terminal The lowest enzyme values were found in patients 1078 Hegnh0j, Hansen, Rannem, S0birk, BjerglundAndersen, Andersen

TABLE iII Duodenal lipase and amylase activities (KU/I) in 143 patients with Crohn's disease likely secreted in the upper part of the intestine. grouped according to localisation and activity ofdisease. Median values (ranges) We found the lowest enzyme values in patients Gut: first published as 10.1136/gut.31.9.1076 on 1 September 1990. Downloaded from Disease activity with extensive disease. All ofthese had disease of the jejunum or ileum proximal to the terminal Mild Moderate Severe ileum, or both. This observation favours the Lipase: hypothesis that decreased secretion of secretory Terminal ileum 979 (324-1732) 556(208-1152)* 635 (184-796)* (n=43) (n= 16) (n= 15) (n= 12) hormones is a factor responsible for decreased Colon 768 (49-1098) 880(300-1412) exocrine pancreatic function in patients with (n= 15) (n= 11) (n=4) Colon+terminal ileum 620(50-1542) 446 (166-1232) 554(5-1060) Crohn's disease. This, however, is probably only (n=59) (n=28) (n=13) (n=18) one of the factors involved, since other investi- Ileum 394 (300-1504) 1413 (730-2095) 655 (126-1388) (n= I1) (n=3) (n=2) (n=6) gators using intravenous stimulation of the Extensive 496(153-1584) 457(65-551) 614(238-774) pancreas, and thus bypassing this factor, also (n= 15) (n=9) (n=3) (n=3) Total 668(49-1732) 556(65-2095) 614(5-1412) found decreased function in several patients with (n= 143) (n=67) (n=33) (n=43) Crohn's disease. I'I Furthermore, we also found Amylase: decreased enzyme activity in some patients with Terminal ileum 271 (101-933) 176 (58-326)* 168 (33-242)* (n=43) (n= 16) (n= 15) (n= 12) colonic disease alone. Colon 167(26-390) 291 (119-727) Pancreatic involvement in patients with (n= 15) (n=-l ) (n=4) Colon+terminal ileum 171(112-488) 136 (47-313) 217 (3-538) Crohn's disease has been hypothesised from (n = 59) (n=28) (n= 13) (n = 18) pathological and functional data, but reports of Ileum 132(74-350) 327 (243-408) 166 (51-264) (n= 1) (n=3) (n = 2) (n= 6) chronic associated with Crohn's Extensive 136(29-447) 91 (26-162) 116(50-194) disease are few and refer to single patients."''2 (n= 15) (n=9) (n=3) (n=3) Total 178 (12-933) 164(26-408) 175 (3-538) The first suggestion of a significant association (n= 143) (n=67) (n=33) (n=43) between pancreatic lesions and chronic inflam- matory bowel disease was reported in 1950 by *Lower than mild activity (p<005) with identical localisation. Ball et al,'3 who found 14% and 53% of macro- scopic and histological lesions, respectively, in a with the most extensive bowel involvement, and postmortem study of patients with chronic compared with values in patients with disease ulcerative . The same group found mild to confined to the terminal ileum these were moderate pancreatic fibrosis at necropsy in 15 of significantly lower. Although the difference 39 patients with Crohn's disease.'4 Axon et al" between enzyme values in other subgroups of observed five cases of histologically proved patients were not significant, the enzyme values chronic inflammatory bowel disease in a series of given in Table I seem to indicate that the extent 59 abnormal pancreatograms, which was too of bowel involvement might be an important many to be considered merely incidental. In the

factor in the impairment of pancreatic function present study none ofthe patients had symptoms http://gut.bmj.com/ in patients with Crohn's disease. The lack of consistent with chronic pancreatitis, but endo- significance in differences between the other scopic retrograde cholangiopancreatography was subgroups might be because the extent ofdisease not performed and histological specimens were involvement in these groups was poorly defined, not available. and therefore probably more variable, so that Crohn's disease is known to be complicated by valid calculations are impossible to perform. an autoimmune involvement of other organs.

For the patient group as a whole no correlation Some investigators have found a high occurrence on September 27, 2021 by guest. Protected copyright. was found between disease activity and enzyme of autoantibodies against pancreatic tissue in values, but for the most uniform group of serum from patients with Crohn's disease,'5 but patients with well defined disease, the patients contradictory reports exist.4 We did not analyse with terminal ileitis, pancreatic function was for autoantibodies, and the question must be significantly lower in the patients with moderate considered open. and severe disease activity compared with In our study no correlation was found between patientswith mild disease. In the other subgroups , judged by albumin concentra- ofpatients no such relation was found. This might tions, general nutritional status, and pancreatic also be caused by the lack ofprecise evaluation of function. disease involvement in the subgroups and there- Our observation is in agreement with a recent fore less uniformity ofpatients in these groups. study.4 In other studies such a correlation has The reason for the decreased pancreatic been suggested,'2 but remains speculative. The function in patients with Crohn's disease is result ofthe Lundh meal test does not necessarily unclear. The following factors might be respons- provide precise information on pancreatic ible: function. The meal test and the cholecystokinin/ (1) Insufficient stimulation of the pancreas test are, however, the only validated, on account of reduced amounts of secretory although indirect, tests for this purpose. We hormones or peptide in the intestinal wall due applied the same test in our control subjects to to or scarring. minimise methodological errors. We cannot (2) Decreased pancreatic function due to exclude a systematic difference in gastric empty- changes in the exocrine pancreas caused by ing as a source oferror, but this is not considered Crohn's disease. plausible as we found no difference between (3) Involvement of the pancreas by autoanti- controls and patients with mild disease activity. bodies against the exocrine pancreatic tissue. In favour of this opinion are the results of (4) Malnutrition resulting from Crohn's previous studies using the cholecystokinin/ disease. secretin test.34 These papers also showed lower Hormones involved in the stimulation of the enzyme activities in patients with Crohn's exocrine pancreas during a meal test are most disease. Pancreaticfunction in Crohn's disease 1079

to 7 Erlandson C, Borgstrom B. Tributyrine as a substrate for We conclude that at least two factors seem determination of lipase activity of pancreatic juice and small be responsible for impaired pancreatic function intestinal content. ScandJ Gastroenterol 5: 293-5.

1970; Gut: first published as 10.1136/gut.31.9.1076 on 1 September 1990. Downloaded from disease: disease activity and localisa- 8 Dahlqvist A. A method for the determination of amylase in in Crohn's intestinal content. ScandJ Lab Invest 1962; 14:145-51. tion or extent of disease. 9 Thaysen EH, Mullertz S, Worning H, Bang HO. Amylase concentration of duodenal aspirates after stimulation of the 1 Dreiling DA. The pancreatic secretion in the malabsorption pancreas by a standard meal. Clinical evaluation in gastro- syndrome and related malnutrition states. J Mt Sinai Hosp intestinal disorders. Gastroenterology 1964; 46: 23-31. 1957; 24: 243-50. 10 Dobrilla G. Regional enteritis and pancreatitis. Gastro- 2 Worning H, Mullertz S, Hess Thaysen E, Bang HO. pH and enterology 1972; 63: 713-4. concentration of pancreatic enzymes in aspirates from the 11 Axon ATR, Ashton MG, Lintott DJ. Chronic pancreatitis human during digestion of a standard meal in and inflammatory bowel disease. Clin Radiol 1979; 30: 179- patients with intestinal disorders. Scand J Gastroenterol 82. 1967;2:81-9. 12 Seyrig JA, Jian R, Modigliani R, et al. Idiopathic pancreatitis 3 Dreiling DA. Studies in pancreatic function. The use of the associated with inflammatory bowel disease. Dig Dis Sci secretin test in the diagnosis of pancreatitis and in the 1985; 30: 1121-6. demonstration of pancreatic insufficiencies in gastrointesti- 13 Ball WP, Baggenstoss AH, Bargen JA. Pancreatic lesions nal disorders. Gastroenterology 1953; 24: 540-55. associated with chronic . Arch Pathol 1950; 4 Angelini G, Cavallini G, Bovo P, et al. Pancreatic function in 50: 347-58. chronic inflammatory bowel disease. IntJr Pancreatol 1988; 14 Chapin LE, Scudamore HH, Baggenstoss AH, Bargen JA. 3:185-93. Regional enteritis associated with visceral changes. Gastro- 5 Lundh G. Pancreatic exocrine function in neoplastic and enterology 1956; 30: 404-15. inflammatory disease; a simple and reliable new test. Gastro- 15 Stocker W, Otte M, Ulrich S, et al. Autoimmunity to enterology 1962; 42: 275-80. pancreatic juice in Crohn's disease. Results of an autoanti- 6 Harvey RF, Bradshaw JM. A simple index of Crohn's disease body screening in patients with chronic inflammatory bowel activity. Lancet 1980; i: 514. disease. ScandJ Gastroenterol 1987; 22 (suppl 139): 41-52. http://gut.bmj.com/ on September 27, 2021 by guest. Protected copyright.