Br. J. Surg. Vol. 62 (1975) 309-312
The cause and treatment of postvagotomy diarrhoea J O H N R . C O N D O N , V. R O B I N S O N , M. I. S U L E M A N , V. S. F A N A N D M. D. M c K E O W N * SUMMARY
Six patients with watery postvagotomy diarrhoea and 4 patients with intractable diabetic diarrhoea were treated with cholestyramine. The diarrhoea responded completely to therapy, during which the stool content of water, sodium, potassium and magnesium fell. Two of the patients, while receiving cholestyramine, had a rise in faecal fat, but this was not accompanied by diarrhoea, the stools being well formed. Evidence is presented which suggests that the mechanism of diarrhoea is similar in both disorders and is due to division or neuropathy of the hepatic fibres o j the vagus nerve, This results in distension of the gallbladder, contraction of which expels increasedquantities of bile salts, which swamp the reabsorptive capacity of the small intestine and induce diarrhoea by direct action on the colon. I f this theory is valid, cholecystectomy, by preventing large quantities of concentrated bile salts fvom suddenly entering the gut, may result in improvement of postvagotomy and diabetic diarrhoea. VAGOTOMYaccompanied by pyloroplasty or gastrojejunostomy is a frequently performed surgical procedure. It is sometimes complicated by diarrhoea which may respond to symptomatic medication or special attention to diet. Failure to respond to these simple measures indicates the presence of intractable diarrhoea, a most difficult long term management problem. A series is reported of 6 postvagotomy patients and 4 diabetic patients with intractable diarrhoea which responded to cholestyramine, an anion exchange resin which binds bile salts in the gut. Our evidence suggests that the mechanism of diarrhoea is similar in both disorders, and a hypothesis is advanced to account for the change in bowel habit which may follow vagotomy. Case reports Case 1 : A 58-year-old man had a truncal vagotomy and pyloroplasty in 1968 because of chronic duodenal ulceration. Division of the vagus was confirmed histologically and a postoperative insulin test showed no response to hypoglycaemia (blood sugar 21 mg/100 ml). Severe watery diarrhoea developed postoperatively but was controlled by codeine phosphate. However, the diarrhoea became progressively more severe and persistent, with 6-7 unformed stools daily, despite continuation of codeine therapy. Codeine phosphate was discontinued for 10 days before treatment with cholestyramine (4 g twice daily) was started. The diarrhoea stopped during cholestyramine administration but recurred when the compound was stopped; symptoms were recontrolled when therapy was restarted. Case 2: A 57-year-old man had a partial gastrectomy in 1969 for a gastric ulcer. Eight months later he had a Heller operation for achalasia of the cardia, but 6 months postoperatively progressive dysphagia again developed. The stomach remnant and lower oesophagus were removed and a Roux-en-Y
oesophagojejunostomy was performed. During this procedure a total vagotomy was carried out. Postoperatively the patient developed severe watery diarrhoea, 6-10 stools daily, there being little response of the diarrhoea to symptomatic treatment which was continued for 3 years. When cholestyramine (4 g thrice daily) was given the diarrhoea stopped, stools became well formed and their number fell to 1-2 per day. Cases 3, 4, 5, and 6: Four subjects all had diarrhoea following vagotomy and pyloroplasty for chronic duodenal ulceration. The patients had 5-8 watery stools daily and there was poor response to symptomatic treatment in all the cases. Bowel habit and faecal appearance returned to normal in every patient when cholestyramine (4 g three times daily) was given. Diarrhoea recurred when cholestyramine was withdrawn but was recontrolled when the compound was restarted. Case 7 : A male aged 37 years had been a diabetic for 24 years. There had been evidence of peripheral neuropathy for 7 years and diabetic retinopathy for 5 years. Diarrhoea had been present for 16 years and was at first intermittent, lasting 3-5 days. The diarrhoea became progressively worse and a t times there were 15-20 bowel actions per day, although 10 stools daily were the average. Investigations o f the gastro-intestinal tract including three barium enema examinations were negative. There was no response to treatment with pancreatic extracts, codeine phosphate, antispasmodics, antibiotics, diphenoxylate and methylcellulose. Oral administration of cholestyramine (4 g thrice and later twice daily) resulted in a reduction of the frequency of stools to 1-2 daily, with ‘solidification’ of the faeces. Diarrhoea recurred when cholestyramine was withdrawn for two separate periods of time, but on both occasions the patient responded when cholestyramine was restarted. Case 8 : A 52-year-old man had suffered from diabetes mellitus for 25 years. There was marked evidence of peripheral neuropathy and diabetic retinopathy. Diarrhoea unresponsive to conventional therapy occurred 6-8 times daily. Cholestyramine ( 4 g twice daily) resulted in a reduction in bowel actions to once daily, the stools being well formed. Withdrawal of cholestyramine resulted in recurrence of diarrhoea until the drug was re-introduced. Cases 9 and 10: Two diabetic subjects with evidence of peripheral neuropathy had had diarrhoea for 4 and 10 years respectively. There had been only limited control of diarrhoea with codeine phosphate. Cholestyramine (4 g once daily and later 4 g twice daily) reduced bowel movements to one per day.
Table Z shows the faecal fat, sodium, potassium, magnesium and fluid contents of the stools in Cases 1 , 2 and 3 when they were and were not treated with cholestyramine. In all the patients cholestyramine was suspended in water and drunk with the main meals.
Discussion We have reported 6 postvagotomy and 4 diabetic subjects with intractable diarrhoea which responded to treatment with cholestyramine. Ayulo (1972) has also reported that patients with postvagotomy diarrhoea respond to oral cholestyramine therapy.
* St Mary’s Hospital, Eastbourne. 309
John R. Condon et al. Table I: STOOL CONSTITUENTS AND BOWEL HABIT IN 3 PATIENTS WITH POSTVAGOTOMY DIARRHOEA BEFORE AND AFTER TREATMENT Case 1 Case 2 Case 3 Reduction Before Daily dosage of cholestyramine (g) Daily stool frequency Total stool weight (g) Dry stool weight (g) Fluid content of stool (g) Stool fat (9) Stool sodium (mEq) Stool potassium (mEq) Stool magnesium (mEq)
After
(7%
Reduction Before
After
(%)
0
8
-
0
12
-
6-7
1 560 143 417 15 2
84 56
6-10 2277 121 2156 34.5 247 142 32
1-2 843 177 666 70.5 18 94 12
80 63 46 * 69 104* 92 33 62
1256 150 1102 22 25 33 -
5
11
52 31 92 66
-
-
Reduction Before
After
(79
0
12
-
7-8 1132 153 979 29.2 70 57 52
1-2 931 222 709 66.0 46 77 36
75 18
45 * 28 126* 43 35 * 31
Stool constituents based on 5-day collection. * Percentage increase.
Postvagotomy and diabetic diarrhoea have many clinical features in common; in both disorders bowel movements are frequent and defaecation may occur without warning, the faeces may be watery in consistency and there may be steatorrhoea, delayed gastric emptying and disordered transit through the small intestine (Wruble and Kalser, 1964). Complete division of the vagi results in enlargement of the gallbladder, which almost doubles its fasting volume and contracts poorly (Johnson and Boyden, 1952; Rudick and Hutchison, 1964; Inberg and Vuorio, 1969). Likewise diabetics are known to have large volume, poorly contracting gallbladders (Gitelson et al., 1963). A high incidence of gallstones is found following vagotomy (Reynolds, 1962; Nielsen, 1964) and also in diabetic subjects (Lieber, 1952). Furthermore, in both disorders examination of the gut has failed to reveal pathological changes to account for the diarrhoea, which is particularly resistant to therapy. The almost invariable association of diabetic diarrhoea with visceral and peripheral neuropathy is highly suggestive of a neuropathy involving the vagus nerve, so that the mechanism of diabetic diarrhoea may be similar to that of diarrhoea following vagotomy. The incidence of diarrhoea following truncal vagotomy can be reduced from 25 per cent to only 2 per cent by selective vagotomy, when the coeliac and hepatic branches of the vagus nerve are preserved (Johnston et al., 1972). These vagal branches supply the biliary system and pancreas. A disturbance of pancreatic exocrine secretion resulting in diarrhoea is, however, unlikely as neither diabetic nor postvagotomy subjects with diarrhoea usually respond to therapy with pancreatic extracts. The fact that our patients with postvagotomy and diabetic diarrhoea responded to cholestyramine (an anion exchange resin which binds with bile acids) strongly suggests that in these disorders excess quantities of bile are reaching the colon. Increased quantities of bile acids have, in fact, been demonstrated in the faeces in postvagotomy patients (Allan et al., 1973a; Gerskowitch et al., 1973), and when individual bile acids are measured there is a highly significant increase in the excretion of chenodeoxycholic acid (Allan et al., 1973b). Saburov (1961) has reported 310
increased bile flow in experimental animals when hydrochloric acid was introduced into the duodenum following section of the vagus. Secretin and other hormones may be involved in the stimulation of bile flow demonstrated by Saburov (1961). There is, however, no general agreement as to the effect of vagotomy, since Baldwin et al. (1966) have demonstrated decreased bile flow following vagotomy. Bile salt concentration is increased following vagotomy (Smith et al., 1973), and although the gallbladder usually contracts within 30 minutes of eating, resulting in a sharp rise in duodenal bile salt concentration (Sjovall, 1959), the concentration of bile salts in the postprandial aspirate is low following vagotomy (Fields and Duthie, 1965). These facts suggest that gallbladder contraction is not in phase with gastric emptying, thus resulting in poor admixture of bile and chyle. Not only may there be poor admixture of bile and chyle but if the overall volume of the gallbladder is approximately twice normal (even in the presence of a mild defect of emptying) increased quantities of bile would pass into the gut. A high concentration of bile salts entering the small gut at inappropriate times could swamp the reabsorptive capacity of the ileum and enter the colon, partly undergoing deconjugation and resulting in diarrhoea (Sobotka, 1937). The probable existence of a limited reabsorptive capacity of the small intestine for bile salts can easily be demonstrated by feeding pure bile salts, when after a few days, severe diarrhoea results (Hardison and Rosenberg, 1967; Hofmann, 1968). Indeed, if free and conjugated bile salts are perfused through the colon, there is inhibition of absorption of sodium chloride, potassium and water (Mekhjian et al., 1971) and disturbances of colonic activity (Galapeaux et al., 1938), all features contributing to the resultant diarrhoea. Wingate et al. (1973) have also shown that conjugated dihydroxy bile acids inhibit electrolyte and fluid absorption and at high concentration evoke secretion of isotonic fluid in the jejunum. It has been suggested that this also contributes to the diarrhoea induced by bile salts (Low-Beer, 1973). The only diarrhoea of known aetiology which responds tocholestyramineis that which occurs following resection or disease of the terminal ileum (Hardison
Postvagotomy diarrhoea and Rosenberg, 1967; Van Deest et al., 1968), when there is malabsorption of bile salts which enter the colon in excess (Row, 1967; Stanley and Nemchausky, 1967 ; Meihoff and Kern, 1968). Chenodeoxycholic acid is thought to play a major role in diarrhoea due to ileal resection (Poley et al., 1972), and Allan et al. (1973a b,) showed that increased faecal concentrations of this bile acid occur in patients with postvagotomy diarrhoea. These findings, taken in conjunction with the fact that patients with diarrhoea due to ulcerative colitis and idiopathic tropical sprue fail to respond to cholestyramine therapy (McCloy and Hofmann, 1971 ; Miettinen, 1971 ;Condon, 1973), are highly suggestive that postvagotomy diarrhoea is due to excess quantities of bile salts entering the colon. Furthermore, oral administration of cholestyramine, by binding bile salts and thus preventing their effects on the gut, resulted in a marked fall in fluid contents of the stool in the present series (Table I ) . The fall in faecal fluid was associated with a decrease in faecal sodium, potassium and magnesium, except in Case 3, in whom stool potassium rose. In 1 patient the faecal fat fell, whereas in 2 other subjects (Cases 2 and 3) steatorrhoea occurred during cholestyramine therapy. Cholestyramine is known to induce steatorrhoea by binding bile salts, and it is of interest that the 2 patients receiving 12 g of the resin daily developed steatorrhoea (though not diarrhoea), whereas the patient taking 8 g of cholestyramine daily had a fall in faecal fat. The optimum dose of cholestyramine should therefore control diarrhoea and result in a fall of faecal fluid, fat, sodium, potassium and magnesium. Furtherexperience of the use of the resin in our patients suggests that a dose of 2 g of cholestyramine twice to four times per day controls diarrhoea without inducing steatorrhoea. We therefore propose that intractable diarrhoea in postvagotomy and diabetic subjects is due to surgical resection or neuropathy of vagal fibres which control gallbladder function. Increased quantities of bile salts are released into the small intestine, swamping its reabsorptive capacity. Diarrhoea is induced by a direct action of bile acids on the colon. The effect of bile salts, which inhibit intestinal absorption of water and electrolytes, can be counteracted by treatment with cholestyramine. If this theory is correct, then cholecystectomy, by preventing large quantities of concentrated bile salts from suddenly entering the gut, may result in improvement of postvagotomy and diabetic diarrhoea.
Acknowledgement V. Robinson was supported by a research grant from the South East Metropolitan Regional Hospital Board. References
v. P. and RUSSELL R . I. (1973a) A study of the role of bile acids in the pathogenesis of post vagotorny diarrhoea. Gut 14, 423424. ALLAN J. G., GERSKOWITCH v. P., and RUSSELL R . I. (1973b) Bile acids and post vagotomy diarrhoea. Br. Med. J. 4. 741. ALLAN J. G., GERSKOWITCH
(1972) Cholestyramine in post vagotomy syndrome. Am. J. Gastroenterol. 57,207-226.
AYULO J. A.
BALDWIN J., HEER W. F., ALBO R . , PELOSO O., RUBY L. and SILEM w . (1966) The effect of vagus nerve
stimulation on hepatic secretion of bile in human subjects. Am. J . Surg. 111, 66-69. CONDON J. R . (1973) Personal observation. FIELDS M. and DUTHIE H. L. (1965) Effect of vagotomy on intraluminal digestion of fat in man. Gut 6 , 301-310. GALAPEAUX E. A., TEMPLETON E. D. and BORKON E. L. (1938) The influence of bile on the motility of the dog’s colon. J. Physiol. 121, 130-136. GERSKOWITCHv. P., ALLAN J. G. and RUSSELL R . I. (1973) Increased secretion of bile acids in post-vagotomy diarrhoea. Br. J. Surg. 60, 912. GITELSON s., SCHWARTS A., FRAENKEL M. and CHOWERS I. (1963) Gall bladder dysfunction in diabetes mellitus. The diabetic neurogenic gall bladder. Diabetes 12, 308-312. HARDISON w . G. M. and ROSENBERG I. H. (1967) Bile salt deficiency in steatorrhoea following resection of the ileum and proximal colon. New Engl. J. Med. 277, 337-342. HOFFMAN A. F. (1968) Function of bile in the alimentary canal. In : CODE c . F. (ed.) Handbook of Physiology, sect. 6, vol. 5. Washington, American Physiology Society, pp. 2507-2533. INBERG M. v. and VUORIO M. (1969) Human gall bladder function after selective gastric and total abdominal vagotorny. Acta Chir. Scand. 135, 625-633. JOHNSON F. E. and BOYDEN E. A. (1952) The effect of double vagotomy on the motor activity of the human gall bladder. Surgery 32, 591-601. JOHNSTON D . , HUMPHREY C. S., WALKER B. E., PULVERTAFT C. N . and GOLIGHER J. C. (1972) VagOtOmy
without diarrhoea. Br. Med, J. 3, 788-790. (1952) The incidence of gall stones and their correlation with other diseases. Ann. Surg. 135, 394-405. LOW-BEER T. s. (1973) Diabetic diarrhoea. Br. Med. J. 4, 741. MCCLOY R . M. and HOFMANN A . F. (1971) Tropical diarrhoea in Vietnam-a controlled study of cholestyramine therapy. New Engl. J. Med. 284, 139-140. MEIHOFF w . E. and KERN F. (1968) Bile salt malabsorption in regional ileitis, ileal resection and mannitol induced diarrhoea. J. Clin. Invest. 47, 261-267. MEKHJIAN H. s., PHILLIPS s. F. and HOFMANN A. F. (1971) Colonic secretion of water and electrolytes induced by bile acids: perfusion studies in man. J. Clin. Invest. 50, 1569-1577. MIETTINEN T. A. (1971) The role of bile salts in diarrhoea of patients with ulcerative colitis. Gut 12,633-635. NIELSEN J. R . (1964) The development of cholelithiasis following vagotomy. Am. J. Digest. Dis. 9, 506508. POLEY J . R . . HOFMANN A. F. and AMMON H. v. (1972) quoted by HOFMAN A. F. (1972) Bile acid malabsorption caused by ileal resection. Arch. Intern. Med. 130. 597-605. LIEBER M. M.
311
John R. Condon et al.
( 1 9 6 2 ) cited by GRIFFITH c. A. (1962) Selective gastric vagotomy. West J. Suug. 70, 175-180. ROW G. G. ( I 967) Control of tenesmus and diarrhoea by cholestyramine administration. Gastroenterology 53, 1 0 0 6 . RUDICK J. and HUTCHINSON .I s. F. ( 1 9 6 4 ) Effect of vagal nerve section on the biliary system. Lancet 1, 579-581. SABUROV G. E. ( 1 9 6 1 ) The effect of vagotomy on the bile secretion function of the liver. Sechenou Physiol. J. USSR 47, 685-692. SJOVALL J. ( 1 9 5 9 ) On the concentration of bile acids in human intestine during absorption. Acta Physiol. Scand. 46,339-345. REYNOLDS R. M.
312
c., MACKAY c. and MCALLISTER R. A. ( 1 9 7 3 ) The effect of vagotomy and drainage on the composition of bile. Scott. Med. J . 18, 65-66. SOBOTKA H. H. ( 1 9 3 7 ) Physiological Chemistry of Bile. London, Bailliere, Tindall & Cox. STANLEY M. M. and NEMCHAUSKY B. ( 1 9 6 7 ) Fecal C14 bile acid excretion in normal subjects and patients with steroid wasting syndromes secondary to ileal dysfunction. J. Lab. Clin. Med. 70, 627-639. SMITH D.
VAN DEEST B. W . , FORDTRAN J. S., MORAWSKI S. G., WILSON J. D. ( 1 9 6 8 ) Bile salt and micellar fat
concentration in proximal small bowel contents of ileectomy patients. J . Clin. Invest. 47, 13141324.
The cause and treatment of postvagotomy diarrhoea J O H N R . C O N D O N , V. R O B I N S O N , M. I. S U L E M A N , V. S. F A N A N D M. D. M c K E O W N * SUMMARY
Six patients with watery postvagotomy diarrhoea and 4 patients with intractable diabetic diarrhoea were treated with cholestyramine. The diarrhoea responded completely to therapy, during which the stool content of water, sodium, potassium and magnesium fell. Two of the patients, while receiving cholestyramine, had a rise in faecal fat, but this was not accompanied by diarrhoea, the stools being well formed. Evidence is presented which suggests that the mechanism of diarrhoea is similar in both disorders and is due to division or neuropathy of the hepatic fibres o j the vagus nerve, This results in distension of the gallbladder, contraction of which expels increasedquantities of bile salts, which swamp the reabsorptive capacity of the small intestine and induce diarrhoea by direct action on the colon. I f this theory is valid, cholecystectomy, by preventing large quantities of concentrated bile salts fvom suddenly entering the gut, may result in improvement of postvagotomy and diabetic diarrhoea. VAGOTOMYaccompanied by pyloroplasty or gastrojejunostomy is a frequently performed surgical procedure. It is sometimes complicated by diarrhoea which may respond to symptomatic medication or special attention to diet. Failure to respond to these simple measures indicates the presence of intractable diarrhoea, a most difficult long term management problem. A series is reported of 6 postvagotomy patients and 4 diabetic patients with intractable diarrhoea which responded to cholestyramine, an anion exchange resin which binds bile salts in the gut. Our evidence suggests that the mechanism of diarrhoea is similar in both disorders, and a hypothesis is advanced to account for the change in bowel habit which may follow vagotomy. Case reports Case 1 : A 58-year-old man had a truncal vagotomy and pyloroplasty in 1968 because of chronic duodenal ulceration. Division of the vagus was confirmed histologically and a postoperative insulin test showed no response to hypoglycaemia (blood sugar 21 mg/100 ml). Severe watery diarrhoea developed postoperatively but was controlled by codeine phosphate. However, the diarrhoea became progressively more severe and persistent, with 6-7 unformed stools daily, despite continuation of codeine therapy. Codeine phosphate was discontinued for 10 days before treatment with cholestyramine (4 g twice daily) was started. The diarrhoea stopped during cholestyramine administration but recurred when the compound was stopped; symptoms were recontrolled when therapy was restarted. Case 2: A 57-year-old man had a partial gastrectomy in 1969 for a gastric ulcer. Eight months later he had a Heller operation for achalasia of the cardia, but 6 months postoperatively progressive dysphagia again developed. The stomach remnant and lower oesophagus were removed and a Roux-en-Y
oesophagojejunostomy was performed. During this procedure a total vagotomy was carried out. Postoperatively the patient developed severe watery diarrhoea, 6-10 stools daily, there being little response of the diarrhoea to symptomatic treatment which was continued for 3 years. When cholestyramine (4 g thrice daily) was given the diarrhoea stopped, stools became well formed and their number fell to 1-2 per day. Cases 3, 4, 5, and 6: Four subjects all had diarrhoea following vagotomy and pyloroplasty for chronic duodenal ulceration. The patients had 5-8 watery stools daily and there was poor response to symptomatic treatment in all the cases. Bowel habit and faecal appearance returned to normal in every patient when cholestyramine (4 g three times daily) was given. Diarrhoea recurred when cholestyramine was withdrawn but was recontrolled when the compound was restarted. Case 7 : A male aged 37 years had been a diabetic for 24 years. There had been evidence of peripheral neuropathy for 7 years and diabetic retinopathy for 5 years. Diarrhoea had been present for 16 years and was at first intermittent, lasting 3-5 days. The diarrhoea became progressively worse and a t times there were 15-20 bowel actions per day, although 10 stools daily were the average. Investigations o f the gastro-intestinal tract including three barium enema examinations were negative. There was no response to treatment with pancreatic extracts, codeine phosphate, antispasmodics, antibiotics, diphenoxylate and methylcellulose. Oral administration of cholestyramine (4 g thrice and later twice daily) resulted in a reduction of the frequency of stools to 1-2 daily, with ‘solidification’ of the faeces. Diarrhoea recurred when cholestyramine was withdrawn for two separate periods of time, but on both occasions the patient responded when cholestyramine was restarted. Case 8 : A 52-year-old man had suffered from diabetes mellitus for 25 years. There was marked evidence of peripheral neuropathy and diabetic retinopathy. Diarrhoea unresponsive to conventional therapy occurred 6-8 times daily. Cholestyramine ( 4 g twice daily) resulted in a reduction in bowel actions to once daily, the stools being well formed. Withdrawal of cholestyramine resulted in recurrence of diarrhoea until the drug was re-introduced. Cases 9 and 10: Two diabetic subjects with evidence of peripheral neuropathy had had diarrhoea for 4 and 10 years respectively. There had been only limited control of diarrhoea with codeine phosphate. Cholestyramine (4 g once daily and later 4 g twice daily) reduced bowel movements to one per day.
Table Z shows the faecal fat, sodium, potassium, magnesium and fluid contents of the stools in Cases 1 , 2 and 3 when they were and were not treated with cholestyramine. In all the patients cholestyramine was suspended in water and drunk with the main meals.
Discussion We have reported 6 postvagotomy and 4 diabetic subjects with intractable diarrhoea which responded to treatment with cholestyramine. Ayulo (1972) has also reported that patients with postvagotomy diarrhoea respond to oral cholestyramine therapy.
* St Mary’s Hospital, Eastbourne. 309
John R. Condon et al. Table I: STOOL CONSTITUENTS AND BOWEL HABIT IN 3 PATIENTS WITH POSTVAGOTOMY DIARRHOEA BEFORE AND AFTER TREATMENT Case 1 Case 2 Case 3 Reduction Before Daily dosage of cholestyramine (g) Daily stool frequency Total stool weight (g) Dry stool weight (g) Fluid content of stool (g) Stool fat (9) Stool sodium (mEq) Stool potassium (mEq) Stool magnesium (mEq)
After
(7%
Reduction Before
After
(%)
0
8
-
0
12
-
6-7
1 560 143 417 15 2
84 56
6-10 2277 121 2156 34.5 247 142 32
1-2 843 177 666 70.5 18 94 12
80 63 46 * 69 104* 92 33 62
1256 150 1102 22 25 33 -
5
11
52 31 92 66
-
-
Reduction Before
After
(79
0
12
-
7-8 1132 153 979 29.2 70 57 52
1-2 931 222 709 66.0 46 77 36
75 18
45 * 28 126* 43 35 * 31
Stool constituents based on 5-day collection. * Percentage increase.
Postvagotomy and diabetic diarrhoea have many clinical features in common; in both disorders bowel movements are frequent and defaecation may occur without warning, the faeces may be watery in consistency and there may be steatorrhoea, delayed gastric emptying and disordered transit through the small intestine (Wruble and Kalser, 1964). Complete division of the vagi results in enlargement of the gallbladder, which almost doubles its fasting volume and contracts poorly (Johnson and Boyden, 1952; Rudick and Hutchison, 1964; Inberg and Vuorio, 1969). Likewise diabetics are known to have large volume, poorly contracting gallbladders (Gitelson et al., 1963). A high incidence of gallstones is found following vagotomy (Reynolds, 1962; Nielsen, 1964) and also in diabetic subjects (Lieber, 1952). Furthermore, in both disorders examination of the gut has failed to reveal pathological changes to account for the diarrhoea, which is particularly resistant to therapy. The almost invariable association of diabetic diarrhoea with visceral and peripheral neuropathy is highly suggestive of a neuropathy involving the vagus nerve, so that the mechanism of diabetic diarrhoea may be similar to that of diarrhoea following vagotomy. The incidence of diarrhoea following truncal vagotomy can be reduced from 25 per cent to only 2 per cent by selective vagotomy, when the coeliac and hepatic branches of the vagus nerve are preserved (Johnston et al., 1972). These vagal branches supply the biliary system and pancreas. A disturbance of pancreatic exocrine secretion resulting in diarrhoea is, however, unlikely as neither diabetic nor postvagotomy subjects with diarrhoea usually respond to therapy with pancreatic extracts. The fact that our patients with postvagotomy and diabetic diarrhoea responded to cholestyramine (an anion exchange resin which binds with bile acids) strongly suggests that in these disorders excess quantities of bile are reaching the colon. Increased quantities of bile acids have, in fact, been demonstrated in the faeces in postvagotomy patients (Allan et al., 1973a; Gerskowitch et al., 1973), and when individual bile acids are measured there is a highly significant increase in the excretion of chenodeoxycholic acid (Allan et al., 1973b). Saburov (1961) has reported 310
increased bile flow in experimental animals when hydrochloric acid was introduced into the duodenum following section of the vagus. Secretin and other hormones may be involved in the stimulation of bile flow demonstrated by Saburov (1961). There is, however, no general agreement as to the effect of vagotomy, since Baldwin et al. (1966) have demonstrated decreased bile flow following vagotomy. Bile salt concentration is increased following vagotomy (Smith et al., 1973), and although the gallbladder usually contracts within 30 minutes of eating, resulting in a sharp rise in duodenal bile salt concentration (Sjovall, 1959), the concentration of bile salts in the postprandial aspirate is low following vagotomy (Fields and Duthie, 1965). These facts suggest that gallbladder contraction is not in phase with gastric emptying, thus resulting in poor admixture of bile and chyle. Not only may there be poor admixture of bile and chyle but if the overall volume of the gallbladder is approximately twice normal (even in the presence of a mild defect of emptying) increased quantities of bile would pass into the gut. A high concentration of bile salts entering the small gut at inappropriate times could swamp the reabsorptive capacity of the ileum and enter the colon, partly undergoing deconjugation and resulting in diarrhoea (Sobotka, 1937). The probable existence of a limited reabsorptive capacity of the small intestine for bile salts can easily be demonstrated by feeding pure bile salts, when after a few days, severe diarrhoea results (Hardison and Rosenberg, 1967; Hofmann, 1968). Indeed, if free and conjugated bile salts are perfused through the colon, there is inhibition of absorption of sodium chloride, potassium and water (Mekhjian et al., 1971) and disturbances of colonic activity (Galapeaux et al., 1938), all features contributing to the resultant diarrhoea. Wingate et al. (1973) have also shown that conjugated dihydroxy bile acids inhibit electrolyte and fluid absorption and at high concentration evoke secretion of isotonic fluid in the jejunum. It has been suggested that this also contributes to the diarrhoea induced by bile salts (Low-Beer, 1973). The only diarrhoea of known aetiology which responds tocholestyramineis that which occurs following resection or disease of the terminal ileum (Hardison
Postvagotomy diarrhoea and Rosenberg, 1967; Van Deest et al., 1968), when there is malabsorption of bile salts which enter the colon in excess (Row, 1967; Stanley and Nemchausky, 1967 ; Meihoff and Kern, 1968). Chenodeoxycholic acid is thought to play a major role in diarrhoea due to ileal resection (Poley et al., 1972), and Allan et al. (1973a b,) showed that increased faecal concentrations of this bile acid occur in patients with postvagotomy diarrhoea. These findings, taken in conjunction with the fact that patients with diarrhoea due to ulcerative colitis and idiopathic tropical sprue fail to respond to cholestyramine therapy (McCloy and Hofmann, 1971 ; Miettinen, 1971 ;Condon, 1973), are highly suggestive that postvagotomy diarrhoea is due to excess quantities of bile salts entering the colon. Furthermore, oral administration of cholestyramine, by binding bile salts and thus preventing their effects on the gut, resulted in a marked fall in fluid contents of the stool in the present series (Table I ) . The fall in faecal fluid was associated with a decrease in faecal sodium, potassium and magnesium, except in Case 3, in whom stool potassium rose. In 1 patient the faecal fat fell, whereas in 2 other subjects (Cases 2 and 3) steatorrhoea occurred during cholestyramine therapy. Cholestyramine is known to induce steatorrhoea by binding bile salts, and it is of interest that the 2 patients receiving 12 g of the resin daily developed steatorrhoea (though not diarrhoea), whereas the patient taking 8 g of cholestyramine daily had a fall in faecal fat. The optimum dose of cholestyramine should therefore control diarrhoea and result in a fall of faecal fluid, fat, sodium, potassium and magnesium. Furtherexperience of the use of the resin in our patients suggests that a dose of 2 g of cholestyramine twice to four times per day controls diarrhoea without inducing steatorrhoea. We therefore propose that intractable diarrhoea in postvagotomy and diabetic subjects is due to surgical resection or neuropathy of vagal fibres which control gallbladder function. Increased quantities of bile salts are released into the small intestine, swamping its reabsorptive capacity. Diarrhoea is induced by a direct action of bile acids on the colon. The effect of bile salts, which inhibit intestinal absorption of water and electrolytes, can be counteracted by treatment with cholestyramine. If this theory is correct, then cholecystectomy, by preventing large quantities of concentrated bile salts from suddenly entering the gut, may result in improvement of postvagotomy and diabetic diarrhoea.
Acknowledgement V. Robinson was supported by a research grant from the South East Metropolitan Regional Hospital Board. References
v. P. and RUSSELL R . I. (1973a) A study of the role of bile acids in the pathogenesis of post vagotorny diarrhoea. Gut 14, 423424. ALLAN J. G., GERSKOWITCH v. P., and RUSSELL R . I. (1973b) Bile acids and post vagotomy diarrhoea. Br. Med. J. 4. 741. ALLAN J. G., GERSKOWITCH
(1972) Cholestyramine in post vagotomy syndrome. Am. J. Gastroenterol. 57,207-226.
AYULO J. A.
BALDWIN J., HEER W. F., ALBO R . , PELOSO O., RUBY L. and SILEM w . (1966) The effect of vagus nerve
stimulation on hepatic secretion of bile in human subjects. Am. J . Surg. 111, 66-69. CONDON J. R . (1973) Personal observation. FIELDS M. and DUTHIE H. L. (1965) Effect of vagotomy on intraluminal digestion of fat in man. Gut 6 , 301-310. GALAPEAUX E. A., TEMPLETON E. D. and BORKON E. L. (1938) The influence of bile on the motility of the dog’s colon. J. Physiol. 121, 130-136. GERSKOWITCHv. P., ALLAN J. G. and RUSSELL R . I. (1973) Increased secretion of bile acids in post-vagotomy diarrhoea. Br. J. Surg. 60, 912. GITELSON s., SCHWARTS A., FRAENKEL M. and CHOWERS I. (1963) Gall bladder dysfunction in diabetes mellitus. The diabetic neurogenic gall bladder. Diabetes 12, 308-312. HARDISON w . G. M. and ROSENBERG I. H. (1967) Bile salt deficiency in steatorrhoea following resection of the ileum and proximal colon. New Engl. J. Med. 277, 337-342. HOFFMAN A. F. (1968) Function of bile in the alimentary canal. In : CODE c . F. (ed.) Handbook of Physiology, sect. 6, vol. 5. Washington, American Physiology Society, pp. 2507-2533. INBERG M. v. and VUORIO M. (1969) Human gall bladder function after selective gastric and total abdominal vagotorny. Acta Chir. Scand. 135, 625-633. JOHNSON F. E. and BOYDEN E. A. (1952) The effect of double vagotomy on the motor activity of the human gall bladder. Surgery 32, 591-601. JOHNSTON D . , HUMPHREY C. S., WALKER B. E., PULVERTAFT C. N . and GOLIGHER J. C. (1972) VagOtOmy
without diarrhoea. Br. Med, J. 3, 788-790. (1952) The incidence of gall stones and their correlation with other diseases. Ann. Surg. 135, 394-405. LOW-BEER T. s. (1973) Diabetic diarrhoea. Br. Med. J. 4, 741. MCCLOY R . M. and HOFMANN A . F. (1971) Tropical diarrhoea in Vietnam-a controlled study of cholestyramine therapy. New Engl. J. Med. 284, 139-140. MEIHOFF w . E. and KERN F. (1968) Bile salt malabsorption in regional ileitis, ileal resection and mannitol induced diarrhoea. J. Clin. Invest. 47, 261-267. MEKHJIAN H. s., PHILLIPS s. F. and HOFMANN A. F. (1971) Colonic secretion of water and electrolytes induced by bile acids: perfusion studies in man. J. Clin. Invest. 50, 1569-1577. MIETTINEN T. A. (1971) The role of bile salts in diarrhoea of patients with ulcerative colitis. Gut 12,633-635. NIELSEN J. R . (1964) The development of cholelithiasis following vagotomy. Am. J. Digest. Dis. 9, 506508. POLEY J . R . . HOFMANN A. F. and AMMON H. v. (1972) quoted by HOFMAN A. F. (1972) Bile acid malabsorption caused by ileal resection. Arch. Intern. Med. 130. 597-605. LIEBER M. M.
311
John R. Condon et al.
( 1 9 6 2 ) cited by GRIFFITH c. A. (1962) Selective gastric vagotomy. West J. Suug. 70, 175-180. ROW G. G. ( I 967) Control of tenesmus and diarrhoea by cholestyramine administration. Gastroenterology 53, 1 0 0 6 . RUDICK J. and HUTCHINSON .I s. F. ( 1 9 6 4 ) Effect of vagal nerve section on the biliary system. Lancet 1, 579-581. SABUROV G. E. ( 1 9 6 1 ) The effect of vagotomy on the bile secretion function of the liver. Sechenou Physiol. J. USSR 47, 685-692. SJOVALL J. ( 1 9 5 9 ) On the concentration of bile acids in human intestine during absorption. Acta Physiol. Scand. 46,339-345. REYNOLDS R. M.
312
c., MACKAY c. and MCALLISTER R. A. ( 1 9 7 3 ) The effect of vagotomy and drainage on the composition of bile. Scott. Med. J . 18, 65-66. SOBOTKA H. H. ( 1 9 3 7 ) Physiological Chemistry of Bile. London, Bailliere, Tindall & Cox. STANLEY M. M. and NEMCHAUSKY B. ( 1 9 6 7 ) Fecal C14 bile acid excretion in normal subjects and patients with steroid wasting syndromes secondary to ileal dysfunction. J. Lab. Clin. Med. 70, 627-639. SMITH D.
VAN DEEST B. W . , FORDTRAN J. S., MORAWSKI S. G., WILSON J. D. ( 1 9 6 8 ) Bile salt and micellar fat
concentration in proximal small bowel contents of ileectomy patients. J . Clin. Invest. 47, 13141324.
