Pathophysiological Responses to Meals in the Zollinger-Ellison Syndrome: 2

Gut: first published as 10.1136/gut.21.2.98 on 1 February 1980. Downloaded from

Gut, 1980, 21, 98-104

Pathophysiological responses to meals in the Zollinger-Ellison syndrome: 2. Gastric emptying and its effect on duodenal function*

J-R MALAGELADAt From the Gastroenterology Unit, Mayo Clinic and Mayo Foundation, Rochester, MN, USA

SUMMARY In this study, wc investigated the relationship between gastric emptying and duodenal events in patients with the Zollinger-Ellison syndrome due to a gastrinoma. Like the inhibitory effect of a meal on gastric secretion (described in a companion paper), postprandial inhibition of gastric emptying reduces fractional gastric emptying rates to normal during the first two hours after a meal. Gastric discharges of content into the duodenum fluctuate considerably, and, in some patients, duodenal acid load and neutralising duodenal secretions appear to be incoordinated. These mechanisms interact in part as a protective system that maintains reasonably normal duodenal homeostasis in most Zollinger-Ellison patients during the early postprandial period. Our data may expJain why clinical evidence of overt-malabsorption is less prevalent and severe in these patients than would be expected from their enormously increased fasting gastric secretory outputs.

Diarrhoea and steatorrhoea are common features of meal and gastric juice, presumably a key patho- of the Zollinger-Ellison syndrome and sometimes genetic factor, has not been quantified. http://gut.bmj.com/ its only clinical manifestations. It is generally In the first paper of this series7 we showed that assumed that they result from the alteration in profound variations of gastric secretory output intraluminal digestion in the upper gut caused by occur postprandially in patients with gastrinoma. gastric emptying of large volumes of highly acidic Thus, it cannot be assumed that the conventional gastric juice. Moreover, several specific disturbances basal acid output or maximal acid output measure- have been identified-among others, acid inactiva- ments also represent the secretory output after a tion of pancreatic enzymes, precipitation of un- meal or, even less so, the postprandial rate of acid on September 29, 2021 by guest. Protected copyright. ionised bile acids, dilution of duodenal contents, delivery into the duodenum. Using intubation and and, possibly, damage to the intestinal mucosa.'-" perfusion techniques developed in our laboratory, In spite of the progress made, our knowledge of we evaluated, in an integrated fashion, the effects of digestive disturbances in the Zollinger-Ellison syn- postprandial gastric emptying on pancreatic, biliary, drome is still fragmentary and incomplete because and intestinal function. Our goal has been to improve (1) only isolated cases of the syndrome have been our understanding of the mechanisms of disturbed studied by any group of investigators, and these intraluminal digestion and diarrhoea in patients case reports do not indicate whether the abnorma- with the Zollinger-Ellison syndrome. lities observed are exceptional or the rule; (2) most of the measurements have been of concentrations Methods of either endogenously secreted or ingested sub- stances without quantification of intraluminal flow PATIENTS or total secretory outputs; and (3) gastric emptying Six patients (two females and four males, aged 13 to 63 years, with a mean of 40-1 years) with the Zollinger-Ellison syndrome were studied. All ful- *Supported in part by Grant AM-6908 from the National Institutes of Health. filled the criteria of fasting gastric hyperchlorhydria, tAddress for reprint requests: DrJ-R Malagelada, Gastro- hypergastrinaemia, and, later, surgically proved enterology Unit, Mayo Clinic, Rochester, MN 55901 gastrinoma. All were normocalcaemic, and none US4. had had gastric surgery before studies were per- Received for publication 10 August 1979 formed. Individual clinical histories and gastric 98 Gut: first published as 10.1136/gut.21.2.98 on 1 February 1980. Downloaded from

Pathophysiological responses to meals in the Zollinger-Ellison syndrome 99 secretory data were reported in the first paper of this into the duodenum at rates (about 600 ml/h, as an series.7 Two patients were restudied after successful average) similar to those of their fasting gastric excision of their gastrinomas (duodenal wall tumour secretory output7, an indlcation that the steady- in both instances) without other surgical procedures. state conditions between gastric secretion and empty- Six healthy volunteers (two women and four men, ing that prevail in the interdigestive period had aged 22 to 63 years, with a mean of 39.7±7.1 returned. At that time, three hours after the meal, years) were similarly studied as controls. intragastric volume was similar in healthy persons in Zollinger-Ellison patients despite quite different QUANTIFICATION OF POSTPRANDIAL GASTRIC emptying rates. This discrepancy is reflected in the EMPTYING, PANCREATIC AND BILIARY fractional gastric emptying rate-that is, the per- SECRETION, AND INTESTINAL FLOW centage of intragastric contents emptied per unit of We used our methods for the quantification of time. Fractional gastric emptying was similar in gastric secretory outputs, gastric emptying of meal Zollinger-Ellison patients and in healthy persons and secretions, pancreatic enzyme and biliary during the first two postprandial hours but during outputs, and total duodenal flow8-10. In the first the third hour it was significantly higher in the former paper of this series7 I have described the pertinent (p<0.05). methodological details. In addition, intestinal samples were analysed for pH, osmolality, trypsin FLOW AT LIGAMENT OF TREITZ AND ITS and lipase activities, and total bile acids by methods RELATIONSHIP TO GASTRIC EMPTYING also previously described.8-'0 In previously reported Intraluminal flow measured at the ligament of studies'" we showed that the presence of the trans- Treitz represents the net volume load (gastric duodenal tube does not significantly alter the rate emptying plus pancreatic, biliary, and duodenal of gastric emptying. Two patients whose tumours secretions minus duodenal absorption) delivered were resected and whose fasting serum gastrin into the jejunum. Flow profiles (Fig. 2) in Zollinger- levels returned to normal after the procedure were Ellison patients were quite different from those in restudied after the fourth postoperative week. healthy controls. The highest flow rates in the healthy persons were reached during the first postprandial Results hour, whereas just the opposite was observed in the Zollinger-Ellison patients. During the second and http://gut.bmj.com/ GASTRIC EMPTYING (Fig. 1) third hours, flow rates were much higher in the During the first postprandial hour, the total volume patients than in the controls (p< 005). of gastric contents (meal and gastric secretion) The influence of gastric emptying on jejunal emptied into the duodenum was similar in healthy volume loads in the Zollinger-Ellison syndrome can controls and in patients with Zollinger-Ellison be appreciated in Fig. 3, where we have compared, syndrome. Thereafter, the volume emptied gradually in each individual, the volume delivered by the decreased in the healthy persons but continued stomach during the three-hour observation period on September 29, 2021 by guest. Protected copyright. increasing in the patients. By the third postprandial with the volume passing the ligament of Treitz hour, gastric juice in the patients was being delivered during the same period. It is apparent that gastric

HEALTH ZOLLINGER-ELLISON

120 _

400 _ Fig. 1 Total volume ofgastric contents empticd into duodenum

80 - after meals, including meal itself (clear area) andaddedgastric mi/lomin 60 - secretion (shaded area). Mean values every 10 minutes aregiven

40 - for healthy controls (left panel) andfor patients with Zollinger- 20 Ellison syndrome (right panel).

0 0 4 2 3 4 0 4 2 3 4 Meal Meal Hours Hours fT Gut: first published as 10.1136/gut.21.2.98 on 1 February 1980. Downloaded from

100 Malagedads

Volume flow at ligament Fig. 2 Intraduodenalflow after meals in healthy persons and in patients with of Treitz, Zollinger-Ellison syndrome. Note different mi/10 min profiles between patients and controls: flow rates informer are highest during /ate postprandialperiod, whereas opposite is true. for healthy individuals.

O 1 2 3 T Meal Hours

emptying is a major determinant of jejunal volume Table 1 Peak postprandial outputs ofpancreatic enzymes load in these patients with Zollinger-Ellison syn- andbile acids drome. This is so because duodenal absorption Peak output* cannot increase enough to compensate for increased Patients fluid emptied and, consequently, flow at the liga- with Trypsin Lipase, Bile acids Z-E (kU/min) (kU/min) (vmol/min) http://gut.bmj.com/ ment of Treitz increases linearly with the absolute rate of gastric emptying. In two patients restudied 1 596 2163 97-1 2 831 3432 81-3 after successful excision of their gastrin-producing 3 1269 1659 79.3 4 1004 4387 53.0 3 5 494 1363 38-1 6 457 1397 45 3 Mean±SE 775 + 131 2400±507 657+-9-6 it 380 2574 48-7

2t 1232 4425 42.5 on September 29, 2021 by guest. Protected copyright. Mean±SE 942 107 2503±283 53.0±5.6 2 0 in 6 Volume of healthy persons gastric contents a O emptied, *Highest three consecutive 10-minute periods. tRepeat studies after excision of gastrin producing tumour. L /3 h

1 .x Table 2 Effect of excision ofgastrinoma on postprandial *0& duodenal events 1 Patient 2 n- - Patient u 3 l 2 Measurement Preop Postop Preop Postop Volume flow at ligament of Treitz, 1 /3 h Lowest duodenal pH 3-2 4-7 2-2 4-7 Fig. 3 Relationship between gastric emptying and Total minutes duodenal pH<4-0 40 0 90 0 intraduodenal after meals in patients with Zollinger- Lowest trypsin concentration flow (U/m1) 25 58 21 119 Ellison syndrome and in healthy controls. Each patient is Lowest lipase concentration identified by a different symbol. A symbol enclosed in a (U/ml) 10 381 1 532 circle indicates postoperative study. Healthy controls are Lowest bile acid concentration Note that (mM) 1.1 40 1.1 3.9 represented by six identical small dots. patients Total minutes duodenal bile detachedfrom control group form linear relationship acidconcentration±2.0mM* 10 0 30 0 between two measurements compared. After surgical removal ofgaYtrinoma, their values approximate control. *Critical micellar concentration. Gut: first published as 10.1136/gut.21.2.98 on 1 February 1980. Downloaded from

Pathophysiological responses to meals in the Zollinger-Ellison syndrome 101

Meal

Acid emptied, meq /10 min

pH Lipase output, U/min x 10-2

Acid emptied. meq / 10 min

pH Lipase output, U/min x 10-2

Acid emptied, meq / 10 min

pH http://gut.bmj.com/ Lipase output. U/min x 10-2

Hews Noers

Fig. 4 Simultaneouspostprandial measurements ofgastric acidemptied (duodenal acidload), on September 29, 2021 by guest. Protected copyright. distal duodenalpH, and total lipase activityfor each ofsix patients with Zollinger-Ellison syndrome. tumours, gastric emptying and intestinal flow re- instances, there was no apparent relationship turned to their normal relationship (Fig. 3). between these two measurements, perhaps a reflec- tion of incoordination between acid delivery and RATES OF ACID DELIVERY INTO DUODENUM duodenal acid-neutralising mechanisms. AND THEIR EFFECTS ON INTRALUMINAL pH, Peak postprandial enzyme outputs (Table 1) PANCREATIC ENZYME ACTIVITY, AND BILE were normal, an indication that both pancreatic ACID CONCENTRATION function and responsiveness of regulatory mecha- In each patient, we measured simultaneously the nisms to the stimulus provided by the meal were postprandial rate of acid delivery into the duodenum normal. Invariably, a decrease in duodenal pH and its effects on pH and lipase activity at the liga- below 4.0 was associated with an abrupt decrease in ment of Treitz (Fig. 4). Almost invariably, we lipase activity (net lipase output). Duration of acid- observed a gradual increase in duodenal acid load inactivation periods varied considerably among during the first 60 to 90 minutes and then wide patients (Fig. 4). In some, the duration was only fluctuations caused by intermittent gushes of secre- 10 to 20 minutes, whereas, in others, inactivation tion emptied by the stomach. Frequently, rapid lasted almost the entire three-hour postprandial increases in duodenal acid load were associated observation period. Concentrations of intraluminal with a marked decrease in duodenal pH. In some trypsin, lipase, and bile acid were also measured Gut: first published as 10.1136/gut.21.2.98 on 1 February 1980. Downloaded from

10,2 Malagelada

100

Trypsin,U/ml

60 40

10 .-L 0. 1 L- 0 1 2 3 0 1 2 3 0 4 1 2 3 Meal Hours Meal Hours Meal Hours

Fig. 5 Concentrations ofintraduodenal trypsin, lipase, and bile acids after meals in healthy persons and in patients with http://gut.bmj.com/ Zollinger-Ellison syndrome. Mean and rangefor healthy controls are illustrated by shaded area. Valuesfor each patient are represented (symbols match those in Fig. 3) and connected by lines. Dotted horizontal line on extreme right panel indicates critical micellar concentration (2 mM).

(Fig. 5) to reveal the added pathogenetic effect of level. In this paper, I describe the events in the excessive dilution of intraduodenal contents by proximal small bowel, relating gastric emptying of gastric secretion. Concentrations of both trypsin meal and secretions to changes in intraluminal pH, and lipase were frequently below the normal ranges concentrations of pancreatic enzyme or bile acid, on September 29, 2021 by guest. Protected copyright. and, during the second and third hours, lipase con- and intestinal flow. centrations often decreased to less than 10% of It is well established that diarrhoea in the Zollin- normal, a critical limit for the development of ger-Ellison syndrome is related to gastric hyper- steatorrhoea.12 Bile acid concentrations were tem- secretion, as continuous aspiration of gastric juice porarily reduced to less than 2 mM (critical micellar and H2-receptor antagonists reduces it or even concentration) in five of the six patients studied abolishes it2 3, 13, 14. Further, many potential patho- (Fig. 5). Correction of the abnormalities described genetic mechanisms of this diarrhoea have been followed the successful excision of a single duodenal enumerated, including dilution of luminal contents, gastrinoma in two of the patients (Table 2). acid-inactivation of pancreatic enzymes, precipitation of bile acids, and alterations of mucosa.'-69 13 Discussion What is not known are the dynamics of the stomach- small bowel interaction, as no simultaneous quanti- In the first paper of this series, I showed that high fication of postprandial gastric emptying and duo- gastric secretory outputs in patients with gastrinoma denal events has previously been reported. were temporarily inhibited by ingestion of a mixed In a recent study in three patients with Zollinger- meal, even though their hypergastrinaemia remained Ellison syndrome, Dubois et al.15 showed a raised unchanged.7 It appeared, therefore, that the meal or fractional gastric emptying that was unchanged by some of its components temporarily counteracted intragastric instillation of a water load. As a group, the stimulatory effect of hypergastrinaemia on gas- the patients reported on here also exhibited a high tric secretion, adjusting it to a more nearly normal fractional gastric emptying but only later in the Gut: first published as 10.1136/gut.21.2.98 on 1 February 1980. Downloaded from

Pathophysiological responses to meals in the Zollinger-Ellison syndrome 103 postprandial period, when most ingested nutrients degree of incoordination appears to exist between had already left the stomach. On the basis of these gastric acid discharges and duodenal neutralising observations, we may conclude that the meal mechanisms. not only temporarily inhibited gastric secretion Can these inhibitory mechanisms of gastric but also corrected abnormally high fractional gastric secretion and emptying offer a perceptible protection emptying. It should not come as a surprise that these against intraduodenal disturbances in Zollinger- meal-induced effects occur during the early post- Ellison syndrome? Our data suggest that, indeed, prandial period, for we have shown that physio- this might be the case, as, during the first post- logically, both the stimulatory and the inhibitory prandial hour (when the bulk of the meal was being effects of the intestinal phase of gastric secretion emptied, as shown in Fig. 1), values for intraluminal and emptying also occur during the first hour after dilution, pH, enzymes, and bile acids were relatively a meal.16, 17 normal in most patients, whereas later in the post- This apparent 'brake' effect of the meal on gastric prandial period, intraduodenal homeostasis was secretion and emptying had repercussions distal to greatly disturbed. Malabsorption is an inconstant the pylorus, because most patients had a surprisingly clinical feature of the Zollinger-Ellison syndrome, gradual increase in gastric acid and volume delivery although, admittedly, accurate balance studies are into the duodenum. In fact, during the first post- not always performed in a clinical setting. In most prandial hour, sharp drops in luminal pH did not reported cases, steatorrhoea is relatively mild'-6, and, occur, concentrations of pancreatic enzyme and as I noted in my previous report,7 Zollinger-Ellison bile acid remained relatively normal, and luminal patients without previous gastric surgery are usually flow at the angle of Treitz actually decreased. well nourished. It is possible that, in many patients, Although fasting duodenal or jejunal pH has been maldigestion and malabsorption affect only the measured in several reported cases1 3' 6, only rarely fraction of the meal that leaves the stomach during has it been measured after a meal. Go et al.6 reported the late postprandial period. This hypothesis would performing postprandial determinations in a patient; be consistent with our data and would explain why as in our patients, the duodenal pH did not begin steatorrhoea is not more prevalent or severe in decreasing sharply until the end of the first post- patients with Zollinger-Ellison syndrome.

prandial hour. It should be emphasised that the Diarrhoea, on the other hand, might be more http://gut.bmj.com/ buffering capacity of the meal itself is too small common and prominent in Zollinger-Ellison (40-1 meq for our meal) to account for this delay patients, because it depends not only on malabsorp- were it not helped by the postprandial decline in tion of ingested fat, so-called fatty acid diarrhoea"8 gastric secretion and emptying that we have shown. but also on massive gastric fluid loads to the gut The magnitude of the reduction in duodenal acid during late postprandial and interdigestive periods. and volume load caused by the meal in Zollinger- It is recognised that, even in normal individuals, Ellison patients is better appreciated by comparing excessive fluid loads may overcome the absorptive duodenal loads during the first postprandial hour capacity of the bowel.19 Estimates of maximal water on September 29, 2021 by guest. Protected copyright. with those during the third hour, when the inhibitory absorption by the human gut oscillate between 10 effect of nutrients (on gastric secretion) had vir- and 15 ml/min20, and these figures are in the range tually disappeared in most patients.7 These varia- of jejunal fluid loads measured in patients with tions during the postprandial period are responsible Zollinger-Ellison syndrome during the late post- for the strikingly different profile of gastric emptying prandial period (Fig. 2). and intestinal flow observed in Zollinger-Ellison In summary, various interacting mechanisms patients and normal individuals. In healthy persons apparently determine intraduodenal conditions pre- (and in patients with idiopathic duodenal ulcer)9" 10, vailing at a given time after a meal in patients with duodenal acid and volume loads increase post- Zollinger-Ellison syndrome. Several pathophysio- prandially, peak usually toward the end of the first logical mechanisms and their deleterious effects on hour, and then decline. In spite of their higher digestion have been identified in the past. In this fasting secretory rate, Zollinger-Ellison patients study, we have focused on less well-recognised have lower duodenal loads during the first post- 'protective mechanisms', which probably help prandial hour than during either fasting (when limit the adverse effects of sustained hypergastri- steady-state conditions between gastric secretion naemia on postprandial gastrointestinal function. and emptying prevail) or the late postprandial period (when the inhibitory effect of the meal on The author thanks Mrs Judith A Duenes, Larry R gastric secretion and emptying has disappeared). Stokes, Ms Brenda A Marben, and Richard Tucker Nevertheless, even in the late postprandial period, for their technical assistance and Ms Gaurdis duodenal acid load fluctuates considerably, and some Grube for her secretarial help. Gut: first published as 10.1136/gut.21.2.98 on 1 February 1980. Downloaded from

104 Malagelada

References "Longstreth GF, Malagelada JR, Go VLW. The gastric response to a tianspyloric duodenal tube. Gut 1975;16: 1Summerskill WHJ. Malabsorption and jejunal ulcera- 777-80. tion due to gastric hypersecretion with pancreatic 12DiMagno EP, Go VLW, Summerskill WHJ. Relations islet-cell hyperplasia. Lancet 1959; 1:120-3. between pancreatic enzyme outputs and malabsorption 2Deleu J, Tytgat H, Van Goidsenhoven GE. Diarihea in severe pancreatic insufficiency. N Engi J Med associated with pancreatic isiet-cell tumours. Am J Dig 1973 ;288:813-5. Dis 1964;9:97-108. "'Donaldson RM Jr, von Eigen PR, Dwight RW. Gastric 3Singleton JW, Kern F Jr, Waddell WR. Diarrhea and hypersecretion, peptic ulceration and islet-cell tumor of pancreatic islet cell tumor: report of a case with a severe the pancreas (the Zollinger-Ellison syndiome): report jejunal mucosal lesion. Gastroenterology 1965;49:197- of a case and review of the literature. N Engl J Med 208. 1957 ;257:965-70. 4Vogel RM, Weinstein LD, Herskovic T, Spiro HM. 4McCarthy DM, Olinger EJ, May RJ, Long BW, Mechanisms of steatorrhea in the Zollinger-Ellison Gardner JD. H2-histamine receptor blocking agents in syndrome. Ann Intern Med 1967;67:816-22. the Zollinger-Ellison syndrcme: experience in seven 1Shimoda SS, Saunders DR, Rubin CE. The Zollinger- cases and implications for long-term therapy. Ann Ellison syndrome with steatorrhea. II. The mechanisms Intern Med 1977 ;87:668-75. of fat and vitamin B12 malabsorption. Gastroenterology "Dubois A, Van Eerdewegh P, Gardner JD. Gastric 1968 ;55 :705-23. emptying and secretion in Zollinger-Ellison syndrome. "Go VLW, Poley JR, Hofmann AF, Summerskill WHJ. J Clin Invest 1977 ;59 :255-63. Disturbances in fat digestion induced by acidic jejunal Clain JE, Malagelada JR, Go VLW, Summeiskill pH due to gastric hypersecretion in man. Gastroenterol- WHJ. Participation of the jejunum and ileum in post- ogy 1970;58:638-46. prandial gastric secretion in man. Gastroenterology 7Malagelada JR. Pathophysiological responses to meals 1977;73:2114. in the Zoflinger-Ellison syndrome: 1. Paradoxical post- 17Clain JE, Go VLW, Malagelada JR. Inhibitory iole of prandial inhibition of gastric secretion. Gut 1978;19: the distal small intestine on the gastric secretory 284-9. response to meals in man. Gastroenterology 1978;74: 'Malagelada JR, Go VLW, DiMagno EP, Summerskill 704-7. WHJ. Interactions between intraluminal bile acids and "Ammon HV, Thomas PJ, Phillips SF. Effects of oleic digestive products on pancreatic and gallbladder func- and ricinoleic acids on net jejunal water and electro-

tion. J Clin Invest 1973 ;52:2160-5. lyte movement: perfusion studies in man. J Clin http://gut.bmj.com/ 9Malagelada JR, Longstreth GF, Summerskill WHJ, Invest 1974;53:374-9. Go VLW: Measurement of gastric functions during 9Levy AG, Benson JW, Hewlett EL, Herdt JR, Dopp- digestion of ordinary solid meals in man. Gastroenterol- man JL, Gordon RS Jr. Saline lavage: a rapid, effective, ogy 1976 ;70:203-10. and acceptable method for cleansing the gastroin- '°Malagelada JR, Longstreth GF, Deering TB, Summer- testinal tract. Gastroenterology 1976; 70:157-61. skill WHJ, Go VLW. Gastric secretion and emptying 20Love AHG, Mitchell TG, Phillips RA. Water and after ordinary meals in duodenal ulcer. Gastroenterol- sodium absorption in the human intestine. J Physio!

ogy 1977;73:989-94. (Lond) 1968 ;195:133-40. on September 29, 2021 by guest. Protected copyright.