Gut 1996; 39: 267-272 267 Gastrointestinal hormones in short bowel syndrome. Peptide YY may be the 'colonic brake'
to gastric emptying Gut: first published as 10.1136/gut.39.2.267 on 1 August 1996. Downloaded from
J M D Nightingale, M A Kamm, J R M van der Sijp, M A Ghatei, S R Bloom, J E Lennard-Jones
Abstract additional major problems in maintaining Background-Short bowel patients with a fluid balance as they lose large volumes of jejunostomy have large volume stomal water, sodium, and magnesium from their outputs, which may in part be due to rapid stoma.' 2 Some of the jejunostomy output is gastric emptying of liquid. Short bowel due to a loss of the normal daily secretions, patients with a preserved colon do not produced in response to food2; some may have such a high stool output and gastric be due to rapid gastric emptying of liquid,3 emptying ofliquid is normal. and some due to increased gastrointestinal Aims-To determine if differences in the secretions, such as gastric acid.4 gastric emptying rate between short bowel In short bowel patients with and without a patients with and without a colon can be colon, the head of the liquid phase of a meal related to gastrointestinal hormone travels more rapidly from the stomach to the changes after a meal. stoma or colon than in normal subjects.3 Subjects-Seven short bowel patients with Following this initial rapid transit, gastric no remaining colon (jejunal length 30- emptying of liquid in those with a colon 160 cm) and six with jejunum in continuity becomes normal, while in those without a colon with a colon (jejunal length 25-75 cm), and it remains fast.3 This finding suggests that the 12 normal subjects. arrival of liquid in the colon, by a neural or Methods-The subjects all consumed a humoral mechanism, is activating a 'colonic 640 kcal meal; blood samples were taken brake' that slows subsequent gastric emptying. for 180 minutes for measurement of To determine if there may be a hormonal gastrointestinal hormones. contribution to differences in gastric emptying Results-Patients with a colon had high rate and gastrointestinal secretions between http://gut.bmj.com/ fasting peptide YY values (median 71 short bowel patients with and without a colon, pmoIIl with a colon; 11 pmoIIl normal we have measured plasma gastrointestinal hor- subjects, p<0005) with a normal post- mone values after a meal in such patients and prandial rise, but those without a colon in normal subjects. had a low fasting (median 7 pmol/l, p=0076) and a reduced postprandial pep- tide YY response (p<0.050). Motilin values Methods on September 25, 2021 by guest. Protected copyright. were high in some patients without a colon. In both patient groups fasting and NORMAL SUBJECTS postprandial gastrin and cholecystokinin Twelve normal healthy volunteers (eight values were high while neurotensin values women and four men) with a mean age of 33 were low. There were no differences years (range 19-53) and taking no medications between patient groups and normal were studied. None had gastrointestinal sym- subjects in enteroglucagon, pancreatic ptoms or previous abdominal surgery. polypeptide, or somatostatin values. Conclusions-Low peptide YY values in short bowel patients without a colon may PATIENTS (TABLE I) cause rapid gastric emptying of liquid. All patients had had a previous resection of St Mark's Hospital, London High values of peptide YY in short bowel their ileum and part of their jejunum. The J M D Nightingale patients with a retained colon may slow patients were clinically stable and well when M A Kamm gastric emptying of liquid and contribute studied at a median time of 48 months (range J R M van der Sijp last resec- J E Lennard-Jones to the 'colonic brake'. 3-181 months) after their intestinal (Gut 1996; 39: 267-272) tion. All were maintaining a stable nutritional, Royal Postgraduate fluid, and electrolyte status and all were taking Medical School, Keywords: short bowel syndrome, peptide hormones, an unrestricted oral diet. All the patients, Hanumersmith peptide YY, satiety. Hospital, London except patient J, were well nourished and had M A Ghatei a body mass index within the normal range. S R Bloom All patients with previous inflammatory bowel Correspondence to: Patients with a short remaining length of small disease had no evidence of current active Dr J M D Nightingale, clinical disease as assessed clinically, biochemically or Department of intestine divide into two common Gastroenterology, Leicester groups: those with a jejunostomy and those radiologically. No patient had had gastric Royal Infirmary, Leicester with jejunum in continuity with a colon.1 surgery. Bowel length was measured at opera- LE1 5WW. have with nutri- tion in nine and radiologically in four Accepted for publication Both patient groups problems 27 February 1996 ent absorption; jejunostomy patients have patients.5 268 Nightingale, Kamm, van der Sijp, Ghatei, Bloom, Lennard-Jones
TABLE I Patient details was eaten or drunk for the following six hours. The patients on home parenteral nutrition had Original Time since Nutritional Age Jejunal cause of last small bowel orfluid stopped their infusion at least two and a half Sex length (cm) short bowel resection (months) supplements hours before the study. Liquid and solid gastric emptying and small Short bowel with no colon Gut: first published as 10.1136/gut.39.2.267 on 1 August 1996. Downloaded from A 32F 30 Crohn's disease 40 HPN bowel transit were determined during this study; B* 32M 50x Ulcerative colitis 68 HPN C 36F 60 Desmoid disease 48 HPN the results have been published previously.3 D 62F 90 Irradiation 3 Oral This study was approved by the ethical E 61M 120 Crohn's disease 181 Oral F 70M 130x Crohn's disease 127 Oral committee of the City and Hackney District G 55M 160 Crohn's disease 146 Oral Health Authority and all subjects gave Short bowel with a colon informed consent. H 19M 25 Volvulus 28 HPN It 25M 25 Volvulus 36 HPN J 41F 70x Adhesions 72 Oral Kt 60M 75x Ischaemia 147 None HORMONE ANALYSIS L4 58F 75 Volvulus 32 Oral Ms 51M 75 Volvulus 23 None Ten ml blood samples for hormone assays were taken into heparinised tubes containing 4000 *Jejunorectal anastomosis. kallikrein inactivator units of aprotinin (0.2 ml tJenunotransverse anastomosis. fIleocaecal valve present. Trasylol). Plasma was separated within 15 §Radiological measurement of bowel length.'7 minutes of sampling and frozen at -20°C until HPN=Home parenteral nutrition. time of assay. The plasma values of neurotensin, motilin, enteroglucagon, pancre- Short bowel with no colon atic polypeptide, gastrin, peptide YY, cholecys- Six patients with an end jejunostomy and one tokinin, and somatostatin were all measured by patient with a jejunorectal anastomosis were radioimmunoassays.6-8The median fasting and included in this group. Four of them had less peak values, and the integrated response (IR) than 100 cm of jejunum remaining, of whom were determined for each hormone. three required longterm home parenteral nutrition. The other four patients required oral electrolyte or nutritional supplements to main- STATISTICAL ANALYSIS tain fluid balance and nutritional status. The Mann-Whitney U test was used to compare each group of patients with normal subjects at all sample times. Short bowel with a colon Six patients with most of their colon in situ were studied. Two had their jejunum anasto- Results
mosed to a short length of terminal ileum (15 http://gut.bmj.com/ cm of ileum in patient L, and 5 cm in patient TOLERANCE OF PROCEDURE M), and had a retained ileocaecal valve and All normal subjects ate the meal within 10 whole colon, two had jejunum anastomosed to minutes with no difficulty. Most of the short the caecum, and two jejunum anastomosed to bowel patients had problems consuming the the proximal transverse colon. Two patients, meal; patient A with no colon could not eat all both with jejunal lengths of25 cm, were receiv- of the pancake even after 20 minutes and left ing longterm home intravenous nutrition. Two 80 g. Three patients with a colon had con- on September 25, 2021 by guest. Protected copyright. others took oral nutritional supplements and siderable difficulty in finishing the meal: one two others required no fluid or nutritional (J) left 63 g after 25 minutes, two others (H supplements. and I) took 12 and 20 minutes to finish the meal. No subject had any problem drinking the orange juice. STUDY DESIGN Antisecretory drugs omeprazole (patient B), octreotide (patient A), and ranitidine (patients PLASMA HORMONE VALUES (TABLE II) C and D) were stopped five, three, and one day respectively prior to the study. The anti- Neurotensin diarrhoeal drugs codeine or loperamide, or both, There was a tendency for fasting and post- (all patients except A, B, H, and I) were stopped prandial values of neurotensin to be lower in the evening before the study; these drugs were both patient groups than controls (Fig 1). In not stopped earlier because of potential drug those without a colon, fasting (Fig 2) and withdrawal symptoms. Women were studied peak neurotensin values correlated with the during the first half of the menstrual cycle. remaining length of jejunum (fasting: r=0-86, A standardised test meal of a pancake and p<0 05, peak: r=0.89, p<0 05). orange juice was chosen; total weight 415 g, energy 640 kcal, carbohydrate 80 g, fat 27 g, protein 18 g, potassium 20 mmol, sodium 6 Motilin mmol, fibre 1 g. After fasting from midnight Motilin values in patients without a colon were the meal was eaten in the morning (about 10 significantly higher than in normal subjects at am) over 10 minutes then the orange juice 0, 90, 120, and 180 minutes (p<0.05) (Fig 3). (195 ml) was drunk. Time 0 is taken as the However those, with the most rapid gastric time of starting to eat the meal. Blood samples emptying of liquid and the shortest remaining were taken at 0, 15, 30, 45, 60, 90, 120, and lengths of jejunum, had motilin values within 180 minutes for hormone assays. Nothing else the normal range. Gastrointestinal hormones in short bowel syndrome. Peptide YY may be the 'colonic brake' to gastric emptying 269
Enteroglucagon Peptide YY Fasting and postprandial enteroglucagon Patients with a colon had high fasting peptide values were not significantly different, in either YY values (median 71 pmol/l) compared with group, from normal subjects. normal subjects (median 11 pmoVl, p<0 005), and peptide YY values were raised at all time points (p<0 005 except 15 minutes when Gut: first published as 10.1136/gut.39.2.267 on 1 August 1996. Downloaded from TABLE II Gastrointestinal hormone results (median and range) p<005). The integrated response was normal Normal No colon Colon (Fig 4). (n=12) (n=7) (n=6) Patients without a colon had low fasting Neurotensin (pmol/l) peptide YY values (median 7 pmol/l, p=0 076) Fasting 46 36 34 and a reduced integrated response (p<005) (15-72) (19-70) (19-49) Peak 74 46* 44 and at all time points peptide YY was less than (29-153) (24-70) (32-84) in normal subjects (p<0 05 except 60 minutes IR 1327 -127 944 (- 1479-11069) (-3723-1914) (- 1215-3963) when p=0.06). Motilin (pmol/l) Fasting 24 86* 19 (5-56) (16-244) (6-68) Peak 51 118 25 Gastrin (6-91) (19-398) (17-115) IR 396 3174 226 Fasting gastrin values were significantly higher (-2475-4859) (-2858-18871) (-1689-6944) in both groups of patients than in normal Enteroglucagon (pmol/l) Fasting 40 30 40 subjects (p<0 05). In those without a colon (16-84) (6-84) (5-162) gastrin values at 30, 60, and 180 minutes and Peak 93 73 134 (29-477) (30-88) (24-250) the integrated response were also significantly IR 1794 3503 3526 higher than in normal subjects (p<005). (-2822-45602) (-4224-4808) (-9424-20422) Peptide YY (pmol/l) Fasting 11 7 71t (6-59) (6-12) (17-179) Peak 33 194 l0ot Cholecystokinin (23-76) (11-23) (58-224) In both patient groups cholecystokinin values IR 1448 261* 3932 (-3656-4728) (-45-1309) (816-10461) (Fig 5) were significantly (p<0 05) higher than Gastrin (pmol/l) in normal subjects at 0 and 180 minutes and Fasting 6 9* 14* (4-11) (5-34) (5-38) additionally in those without a colon at 90 Peak 27 63 34 minutes. (8-78) (6-142) (8-197) IR 1069 2284* 1354 (102-2123) (60-11318) (- 1665-14108) Cholecystokinin (pmol/1) (no colon n=6) Fasting 3-3 4.5t 4.0* Pancreatic polypeptide (2.2-4.2) (3.9-9.4) (3-7-11-5) Fasting and postprandial pancreatic polypep- Peak 6-1 10.8* 9-8 in http://gut.bmj.com/ (4-3-10.3) (3-9-13-6) (4-4-14-6) tide values were not significantly different, IR 339 464 560 either group, from normal subjects. (-65-740) (-60-868) (32-909) Pancreatic polypeptide (pmol/l) (no colon n= 6) Fasting 34 32 46 (12-60) (17-87) (32-88) Peak 167 263 247 Somatostatin (35-579) (49-599) (59-695) Fasting and postprandial somatostatin values IR 17052 26448 24367 (1182-73690) (1862-66484) (-229-91010) were not significantly different, in either group, Somatostatin (pmolA1) from normal subjects. on September 25, 2021 by guest. Protected copyright. Fasting 30 41 43 (18-52) (18-73) (22-70) Peak 52 61 59 (41-75) (31-114) (419-9) Discussion IR 2189 787 1762 (-87-4657) (-3047-11781) (-2878-2384) The plasma values of peptide YY show the most striking difference between short bowel IR units pmol/1/80 min. 8 with normal patients and normal subjects. In those with a *P<0u05,tp<0mo05 p<0.001 compared subjects. preserved colon fasting and postprandial values of peptide YY were higher than normal subjects, while in those without a colon they 80 _ were lower. Peptide YY, which is produced in the ileum and colon,7-10 is released when unabsorbed nutrients, especially fat,7 11 or bile - 60 salts12 reach the terminal ileum and colon. E High plasma concentrations of peptide YY TT T 1 /- T TT have been reported in tropical sprue, chronic dumping syndrome,14 and after ° 40t40 panII 25-5 +t1T1ll q* IIL ,,anrtcreatitis,13ileal resection leaving the colon in situ.15 16 Low values occur in patients who have had a
0 20 - O 20 ~~~~~~~~~~~~~~~colectomy.15Peptide YY slows gastric emptying.17 18 The failure of peptide YY to rise after a meal in short bowel patients without a colon, may explain why the gastric emptying rate for liquid o 0 30 60 90 120 150 180 does not slow. This rapid gastric emptying rate Time (min) may be a factor in causing a large jejunostomy Figure 1: Graphfor normal subjects 0, patients with 0 and without A a colon for output. As liquid enters the colon, a rise in the neurotensin.- - Median ---and interquartile.2 ---- ranges------~ shown.~ - - - - already high fasting peptide YY value occurs, 270 Nightingale, Kamm, van der Sijp, Ghatei, Bloom, Lennard-Jones
and a,s we have shown previously it is at this gastric acid to be secreted20 and less water and time 1that the gastric emptying rate of liquid electrolyte to be absorbed by the small returnIS to normal.3 Thus we suggest that pep- bowel.2' tide YY is responsible for this slowing of subse- The initial fast transit of the head of the quent gastric emptying. Serum concentrations liquid meal, in both patient groups,3 may of peptide YY, similar to those we have reflect loss of neurotensin producing cells in Gut: first published as 10.1136/gut.39.2.267 on 1 August 1996. Downloaded from measuLred in patients with a colon, slowed the ileum.22 The fasting values of neurotensin, gastrimc emptying in normal subjects.'8 It has which slows gastric emptying,23 were low in been suggested that peptide YY is the major both patient groups. In those without a colon horm(Dne responsible for the 'ileal brake'"9 and fasting neurotensin values correlated with now vYe suggest that it is also responsible for residual jejunal length. This finding may the 'colonic brake'. In addition to quickening explain why those with the shortest length of the rate of gastric emptying, low peptide YY remaining jejunum and the lowest neurotensin values in patients without a colon, may values have the most rapid rates of liquid increase jejunostomy output by causing more gastric emptying.3 The values of motilin, which is produced by the upper small intestine24 25 and increases the 70 A rate of were in r < gastric emptying,2628 high o - 0.86 p 0.05 patients without a colon. However, low or E 60 A Q50 normal motilin values were seen in those with- out a colon, who had the shortest remaining o 40 A A lengths of jejunum and the most rapid rates of X30 A liquid gastric emptying. Thus high motilin 0c 20 values are not responsible for rapid early 0 o10 gastric emptying of liquid. I I I I I I I Hormonal changes may affect gastrointesti- 00 20 40 60 80 100 120 140 160 nal secretions and hence the volume of output Jejunal length (cm) from a jejunostomy. Both fasting and post- Figure 2: Graphforpatients without a colon offasting prandial gastrin values were high in both neurote nsin plotted against residualjejunal length. patient groups. While previous drug therapy (for example, with a proton pump inhibitor or H2 antagonist) could contribute to the high gastrin values in those without a colon, this is unlikely. One patient without a colon had stopped taking 20 mg omeprazole daily, five
0 days before the study. This time period should http://gut.bmj.com/ E have allowed serum gastrin concentrations to C. return to pre-treatment values.29 The two 0 patients who had been taking ranitidine, which a,; only increases plasma gastrin values by a small * amount,30 stopped taking it at least 24 hours 0 before the study, when the effect of ranitidine 0 on gastric acid secretion should have ceased.3' on September 25, 2021 by guest. Protected copyright. Removal of the colon alone does not affect gastrin concentrations.32 High gastrin values | . probably... , _reflect the loss of small bowel, which 0 30 60 90 120 150 180 is a major site of gastrin catabolism.33 34 High Time (min) gastrin concentrations have been reported in Figure 3: Graph for normal subjects 0, 1patients with 0 and without A a colonfor motilin. patients with a short bowel and a colon35 36 Median and interquartile ranges shown. *p 2 Nightingale JMD, Lennard-Jones JE, Walker ER, Farthing MJG. Jejunal efflux in short bowel syndrome. Lancet 12 1990; 336: 765-8. 3 Nightingale JMD, van der Sijp JRM, Kamm MA, Walker ER, Mather SJ, Britton KE, et al. Disturbed gastric 10 * emptying in the short bowel syndrome. Evidence for a 'colonic brake'. Gut 1993; 34: 1171-6. 4 Buxton B. Small bowel resection and gastric acid hyper- Gut: first published as 10.1136/gut.39.2.267 on 1 August 1996. Downloaded from 0 secretion. Gut 1974; 15: 229-38. 5 Nightingale JMD, Bartram CI, Lennard-Jones JE. Length of residual small bowel after partial resection: correlation E between radiographic and surgical measurements. 0 GastrointestRadiol 1991; 16: 305-6. cc 6 Bloom SR, Long RG. Radioimmunoassay of gut regulatory 4 peptides. London: W B Saunders, 1982. 7 Adrian TE, Ferri G-L, Bacarese-Hamilton AJ, Fuessl HS, 0 2 _ Polak JM, Bloom SR. Human distribution and release of a putative new gut hormone, Peptide YY. Gastroenterology 2 1985; 89: 1070-7. 8 Jansen JBMJ, Lamers CBHW. Radioimmunoassay of cholecystokinin in human tissue and plasma. Clin Chim 0 Acta 1983; 131: 305-16. 0 30 60 90 120 150 180 9 Ali-Rachedi A, Varmdell IM, Adrian TE, Gapp DA, Van Noorden S, Bloom SR, et al. Peptide YY (PYY) Time (min) immunoreactivity is co-stored with glucagon-related immunoreactants in endocrine cells of the gut and pan- Figure 5: Graph for normal subjects 0, patients with * and without A a colon for creas. Histochemistry 1984; 80: 487-91. cholecystokinin. Median and interquartile ranges shown. *p intragastric acidity and plasma gastrin concentration intake and gastric emptying in man. Physiol Behav 1988; before and during treatment with either ranitidine or 44: 645-9. omeprazole. Aliment Pharmacol Therap 1987; 1: 239-51. 42 Lieverse RJ, Jansen JBMJ, Masclee AAM, Lamers CBHW. 31 Roberts CJC. Clinical pharmocokinetics of ranitidine. Clin Satiety effects of cholecystokinin in humans. Gastro- Pharmacokinet 1984; 9: 211-21. enterology 1994; 106: 1451-4. 32 Kennedy HJ, Sarson DL, Bloom SR, Truelove SC. Gut 43 Porus RL. Epithelial hyperplasia following massive small hormone responses in subjects with a permanent bowel resection in man. Gastroenterology 1965; 48: 753-9. ileostomy. Digestion 1982; 24: 133-6. 44 Weinstein LD, Shoemaker CP, Hersh T, Wright HK. Gut: first published as 10.1136/gut.39.2.267 on 1 August 1996. Downloaded from 33 Temperley JM, Stagg BH, Wyllie JH. Disappearance ofgas- Enhanced intestinal absorption after small bowel resec- trin and pentagastrin in the portal circulation. Gut 1971; tion in man. Arch Surg 1969; 99: 560-2. 12: 372-6. 45 Dowling RH, Booth CC. Functional compensation after 34 Becker HD, Reeder DD, Thompson JC. Extraction of small bowel resection in man: demonstration by direct circulating endogenous gastrin by the small bowel. measurement. Lancet 1966; ii: 146-7. Gastroenterology 1973; 65: 903-6. 46 O'Keefe SJD, Shorter RG, Bennet WM, Haymond MW. 35 Straus E, Gerson CD, Yalow RS. Hypersecretion of gastrin Villous hyperplasia is uncommon in patients with massive associated with the short bowel syndrome. intestinal resection. Gastroenterology 1992; 102: A231. Gastroenterology 1974; 66: 175-80. 47 Gleeson MH, Bloom SR, Polak JM, Henry K, Dowling RH. 36 Williams NS, Evans P, King RFGJ. Gastric acid secretion Endocrine tumour in kidney affecting small bowel struc- and gastrin production in the short bowel syndrome. Gut ture, motility, and absorptive function. Gut 1971; 12: 1985; 26: 914-9. 773-82. 37 Windsor CWO, Fejfar J, Woodward DAK. Gastric secre- 48 Bloom SR. An enteroglucagon tumour. Gut 1972; 13: tion after massive small bowel resection. Gut 1969; 10: 520-3. 779-86. 49 Rudo ND, Rosenberg IH. Chronic glucagon administration 38 Bloom SR, Polak JM. Somatostatin. BMY 1987; 295: enhances intestinal transport in the rat. Proc Soc Exp Biol 288-9. Med 1973; 142: 521-5. 39 Nightingale JMD, Walker ER, Burnham WR, Farthing MJG, 50 Miazza BM, Al-Mukhtar MYT, Salmerson M, Ghatei MA, Lennard-Jones JE. Octreotide (a somatostatin analogue) Felce-Dachez M, Filali A, et al. Hyperenteroglucagon- improves the quality of life in some patients with a short aemia and small intestinal mucosal growth after colonic intestine. Aliment Pharmacol Therap 1989: 3: 367-73. perfusion of glucose in rats. Gut 1985; 26: 518-24. 40 Sturdevant RAL, Goetz H. Cholecystokinin both stimulates 51 de Miguel E, Gomez de Segura IA, Bonet H, Rodriguez- and inhibits human food intake. Nature 1976; 261: 713-5. Montes J A, Mata A. Trophic effects of neurotensin in 41 Muurahainen N, KissileffHR, Derogatis AJ, Pi-Sunyer FX. massive bowel resection in the rat. Dig Dis Sci 1994; 39: Effects of cholecystokinin-octapeptide (CCK-8) on food 59-64. http://gut.bmj.com/ on September 25, 2021 by guest. Protected copyright.