- Email: [email protected]
Duodenogastric reflux in humans: Its relationship to fasting antroduodenal motility and gastric, pancreatic, and biliary secretion
GASTROENTEROLOGY
1981;81:728-31
Duodenogastric Reflux in Humans: Its Relationship to Fasting Antroduodenal Motility and Gastric, Pancreatic, and Biliary Secretion FRANK B. KEANE, EUGENE and JUAN-R. MALAGELADA Gastroenterology
Unit,
Mayo
Clinic
and
P. DIMAGNO, Mayo
Duodenogastric reflux may have pathophysiologic importance, but its mechanism is poorly understood. We propose that duodenogastric reflux and periodic changes in motor and secretory activity of the upper gut during fasting may be related. Therefore we determined the relationships between duodenogastric reflux and interdigestive motor-secretory cycles in a group of 6 healthy individuals, on each of whom we performed three 7-h studies in random order on separate days. In all studies gastric intubation and antral pressure recordings were performed. However, in design 1 we used a slow duodenal perfusion rate of rC]PEG in saline (0.25 mI/min) while in design 3 we used a fast perfusion rate (3.0 ml/min). Duodenal pressures were also recorded during these designs. In design 2 no transpyloric tubes were present. Our study shows that, in humans, fasting duodenogastric reflux of bile and pancreatic juice is cyclic and closely related to the interdigestive migrating motor complex. Reflux is highest during late phase II (when secretory activity is also on the rise) and lowest after phase III. One of the important functions of the migrating motor complex in humans may be to clear the stomach of refluxed duodenal secretions. In humans, reflux of duodenal contents into the stomach occurs sporadically during fasting and after meals (1). Increased duodenogastric reflux has been found in patients with gastric ulcer, esophagitis, and gallstone dyspepsia (2). Bile acids and pancreatic enzymes, normally found in the duodenum, have been Received August 8, 1980. Accepted May 22, 1981. Address requests for reprints to: Juan-R. Malagelada, M.D., Gastroenterology Unit, Mayo Clinic, Rochester, Minnesota 55905. This investigation was supported in part by Grants CA 25064 and AM 26428 from the National Institutes of Health, Public Health Service, Bethesda, Maryland. Dr. Malagelada is the recipient of Research Career Development Award AM 00330 from the National Institutes of Health. 0 1981 by the American Gastroenterological Association OOlS-5085/81/100726-08$02.50
Foundation,
Rochester,
Minnesota
shown experimentally to be capable of damaging the gastric and esophageal mucosa (3-6). However, while the mechanism of duodenogastric reflux may be both physiologically and pathophysiologically important, it is poorly understood (7). Two premises, however, seem valid: first, duodenal contractile activity is the propulsive force for duodenal contents to move in a retrograde direction, and, second, the pylorus must be open, at least partially, for this material to enter the stomach. During fasting, gastrointestinal motility is organized in alternating cycles of activity and quiescence. Characteristic features of the former are its distal propagation from the stomach to the ileum and its association with an increase in gastric, pancreatic, and biliary secretion (8-10). These current concepts of fasting motor and secretory activity of the gut suggest that duodenogastric reflux is also related to interdigestive motor-secretory cycles. Preliminary data from our laboratory have indicated that phase III (intense, propagated activity) of the interdigestive motor cycle may play an antireflux role (11). However, very little work along these lines has been performed to date. In the present study we have attempted to characterize the temporal relationships between motor-secretory cycles and duodenogastric reflux in health. In order to resolve potential technical difficulties and theoretical objections that might arise secondary to the use of gastroduodenal intubation and perfusion, we employed three different experimental designs to quantify motor and secretory activity and duodenogastric reflux in each individual.
Materials, Design
Methods,
and Experimental
Studies were performed on 6 healthy aged 23-53 yr. Informed consent was obtained.
volunteers In random
October 1981
INTERDIGESTIVE
order, three studies were performed on each volunteer on 3 different days. The three studies involved different assemblies of gastrointestinal tubes (Figure 1). After an overnight fast a volunteer was asked to swallow one of these assemblies. After the position of the tubes was confirmed fluoroscopically, the volunteer assumed a comfortable semirecumbent position. The purpose of using three designs in this study was to: (a) evaluate simultaneous gastric and duodenal motor-secretory activity under near physiologic conditions; (b) determine whether a transpyloric tube alters duodenogastric reflux; and (c) determine whether a marked increase in duodenal perfusion rate would change our results. Accordingly, in design 1 very low gastric and duodenal perfusion rates were used, in design 2 we omitted the duodenal tube, and in design 3 duodenal perfusion rates were 12-fold greater (3 ml/min) than in design 1.
Design 1 Study (Slow Duodenal
Perfusion)
Two sump-type triple-lumen polyvinyl tubes were positioned so that the tip of one lay at the ligament of Treitz and the tip of the other was in the dependent part of the stomach (Figure 1). The duodenal tube contained two miniature strain gauge pressure transducers (Millar Instruments, Houston, Tex., Model PC 350), one positioned in the antrum and the other at the ligament of Treitz. One lumen of the duodenal tube was located opposite the ampulla of Vater and was used to perfuse polyethylene-l-2[‘4C]glycol ([Y]PEG) (New England Nuclear, Boston, Mass.), 0.2 &i/ml 154 mM NaCl, at 0.25 ml/min. Perfusion was carried out with a low compliance perfusion system (12) which permitted simultaneous pressure monitoring. A strain gauge (Statham Instruments, Hato Rey, Puerto Rico) attached to a Gould recorder was used to record pressures. The second lumen of the duodenal tube was used to continuously aspirate fluid at the ligament of Treitz. The third lumen acted as an air vent. A fingercot containing 0.75 ml of mercury was attached to the distal end of the tube to facilitate its passage into the duodenum. The gastric tube was also used for perfusion and aspiration. The perfusate contained a second nonabsorbable marker, polyethylene-l-2-[3H]-glycol ([3H]PEG) (New Engxstm
1-3
PERFUSION
RATE
TRANSPYLORIC NO TRANSPYLORIC NO DUODENAL
REFLUX
727
land Nuclear, Boston, Mass.), 0.2 &i/ml 154 mM NaCl. This solution was perfused by means of a syringe pump (Harvard Apparatus, Dover, Mass.) through a port 10 cm proximal to the tip of the gastric tube, at a rate of 0.072 ml/ min. Meanwhile, gastric contents were continuously aspirated from the tip of the gastric tube. Gastric and duodenal aspirates were simultaneously collected every 10 min. Gastric juice was analyzed for pH, titratable acidity, bile acids, [3H]PEG, and [Y]PEG. Duodenal samples were analyzed for pH, bile acids, trypsin, bicarbonate, and [‘*C]PEG.
Design 2 Study (No Duodenal
Tubes)
Two polyvinyl tubes were placed in the stomach. To one of these gastric tubes two miniature strain gauge pressure transducers were attached, 3.5 cm apart, and positioned so as to record antral motility (Figure 1). The second gastric tube was a triple-lumen sump tube. As in the previous study design, [3H]PEG was perfused at 0.072 ml/ min, 10 cm proximal to the tip of this tube. From the tip, gastric juice was sampled at IO-min intervals and analyzed for pH, titratable acidity, bile acids, and [3H]PEG.
Design 3 Study (Fast Duodenal
Perfusion)
A duodenal tube and a gastric tube similar to those used in the design 1 study were used. However, the gastric tube was used only for sampling gastric contents and no marker was perfused into the stomach. The duodenal tube carried two miniature strain gauge pressure transducers positioned 22 cm apart such that one lay in the antrum and the other lay in the second part of the duodenum opposite the ampulla of Vater (Figure 1). Polyethylene glycol (PEG 4000), 5 g/l 154 mM NaCl, heated to 37% was perfused at 3 ml/min through a polyvinyl tube whose opening also lay opposite the ampulla of Vater. Through another fine polyvinyl tube attached to the duodenal tube, also opening at the second portion of the duodenum, 0.5-ml pulses of a second nonabsorbable marker, [3H]PEG (3 pCi/ ml), were injected at intervals and flushed with 1.5 ml saline, a volume equal to the dead space volume of the tube. The interval between injection and subsequent recovery
e-
-JIG-
TRANSPYImORIC TlJBE t.OW DUODENAi
ACTIVITY AND DUODENOGASTRIC
TUBE
PERFUSION
TUBE
FAST DUODENAL PERFUSICN
RATF
Figure 1. Assemblies used to simultaneously measure and validate measurements of gastric, biliary, and pancreatic secretion and antroduodenal motility.
728
GASTROENTEROLOGY
KEANE ET AL.
of this marker in duodenal collections taken every 5 min for 20 min after injection represented the time necessary for duodenal contents to travel from the ampulla of Vater through the duodenum and collecting catheter. Duodenal aspirates were collected continuously at the ligament of Treitz every 10 min and analyzed for pH, bile acids, trypsin, bicarbonate, PEG 4000, and [3H]PEG. Gastric samples taken during each lo-min interval were analyzed for pH, titratable acidity, and bile acids. All of the studies were begun after a 1-h equilibration period and continued for approximately 7 h (average duration for design 1, 415 f 14 min, SEM, n = 6; for design 2, 440 f 15 min; and for design 3, 390 f 11 min). Each Millar transducer was connected to a Gould four-channel recorder (Gould Model 2400, Cleveland, Ohio) via a control unit (Millar Instruments, Houston, Tex., Model TC 100) that provided zero and 100 mmHg calibration at room temperature (thermal stability of sensors = 1 mmHg over room temperature, range 23’-38’C). The lo-min sampling intervals were marked on each motility record to facilitate analysis.
Analytic
Procedures
Using a glass electrode, pH was measured. Titratable acidity of the samples was determined by titration with NaOH. Bile acid concentrations were determined by the steroid dehydrogenase method (13) trypsin by titrimetric assay (14) and bicarbonate by CO, analyzer (Ericsen Instruments, Ossining, N.Y., Model E-109). [l’C]PEG and [3H]PEG were assessed by /3-scintillation counting while PEG 4000 was determined turbidimetrically. Outputs were calculated based on the recovery of the nonabsorbable markers. During design 1, the slow duodenal perfusion study, sample amounts obtained from the ligament of Treitz were frequently insufficient for analytical purposes (70% of the lo-min samples). Antroduodenal motility records were divided into lomin periods corresponding to the sampling intervals. Records were analyzed for the start of phase III motor activity in the second part of the duodenum, phase III being defined as regular duodenal contractions at a frequency not less than 8/min and continuing for at least 3 min. During design 1 and 3 studies, propagation of the activity front (phase III) could be followed from the antrum to the duodenum. In these studies we observed that the end of phase III activity in the antrum corresponded to the start of phase III activity in the second part of the duodenum. Therefore, during design 2 studies, when recordings were only obtained from the antrum, the onset of phase III activity in the second part of the duodenum was defined as the end of phase III in the antrum. The end of phase III in the antrum was identified as the cessation of 3 contractions/min.
Statistical
Methods
For the purpose of analysis, collection periods were grouped in relation to appearance of phase III activity in the second portion of the duodenum. Three lo-min collection periods before and two lo-min periods after the
Vol. Al. No. 4
onset of phase III activity were considered to be associated with the onset of phase III activity while the remain-
ing lo-min collections were considered to be unassociated. Results were averaged across episodes of phase III activity for the 6 patients and this mean value per patient was used in subsequent statistical analysis. Significant differences among individual values for marker reflux, bile acid output, and other secretory parameters during the three IO-min intervals before and after the onset of phase III were determined by using the sign test.
Results Motor Activity During most experiments phase III activity recurred at intervals which were not always regular. One volunteer had less frequent activity fronts than the others-two on one day, one on the other, and none on the third day. During design 1 and design 3 studies, over 90% of the activity fronts were propagated from the antrum to the duodenum. The mean frequency of phase III activity for each volunteer during each experiment was just over 3 with a range varying from 0 to 5. The frequency of phase III activities for each of the experimental designs was very similar (average number of phase III activities was 3.3 + 0.6, SEM, n = 6 for design 1, 3.2 + 0.4 for design 2, and 3.3 + 0.7 for design 3) suggesting that the frequency of activity fronts was not altered by a transpyloric tube or by slow (0.25 ml/ min) or fast (3 ml/min) duodenal perfusion. Secretion Increases in gastric acid output (from design 1 and design 2 studies), and duodenal bile acid, trypsin, and bicarbonate outputs (from design 3 study) were associated with the onset of phase III motor activity (Figure 2). Mean gastric acid output and duodenal bicarbonate outputs were maximal during the IO-min period after the onset of phase III while mean duodenal bile acids and trypsin outputs were higher prior to the onset of phase III activity (Figure 3). Average values for each of the variables 30 min before and after phase III differed significantly (p < 0.05) by the sign test. During the design 3 study the transit time from the ampulla of Vater to the collection site for duodenal samples was measured using injected pulses of [3H]PEG. The phase of motor activity at the time of injection was defined as follows: phase I was a flat tracing reflecting absent duodenal motor activity, phase II reflected irregular contractions, and phase III reflected a burst of regular contractions as previously described. Peak recovery of marker pulses at the ligament of Treitz occurred within 10 min re-
INTERDIGESTIVE ACTIVITY AND DUODENOGASTRlC
1981
October
REFLUX
729
Representative example showing phases of motor activity in association with gastric acid output and duodenal outputs of trypsin and bile acids. Regular recurring fronts. Phase I, white portion; phase II, stippled portion; phase III, solid black portion.
L
I
I
I
1
2
1
-3e--
3--'
Time (hours)
gardless of the motor phase and no statistically significant differences were noted (data not shown). By 30 min after injection of marker pulses a mean of 58% had been recovered after injection during phase I. 59.2% during phase II, and 57.2% during phase III. Duodenogastric
Reflux
Duodenogastric reflux of [“CIPEG was meaduring the design 1 study when the stomach sured was continuously aspirated. Reflux was maximal prior to the onset of the activity front. Following the onset of phase 111activity intragastric [W]PEG was cleared from the stomach (Figure 4). Average
Bile acid out ut (pmol/min P
Gastrcc acid output (mEq/mm) 4 $ z -k *SE * = p
01 HCOj
,
’
[Y]PEG reflux 30 min before and after phase III differed significantly by the sign test (p c 0.05). Mean intragastric bile acid concentration, measured for each volunteer during the three studies, was maximal prior to the onset of phase III activity (Figure 5). There were no consistent differences in mean gastric bile acid concentrations among the three study designs when the means from each design were compared (data not shown). In 4 individuals who had three or more episodes of phase III during day 3, we compared actual intragastric bile acid content (determined from gastric samples by multiplying intragastric bile concentration by gastric volume) with the predicted bile acid reflux (deter-
05
15and+25min
I
1
I
I
I
0
,_.-__L
VS + 5, + 15 + 25 min
...._..-L__i
output
vD”“mi”)
I:
F
T
16
tSE 05vs+
T **=p< + =p’
15and+25mm
05 vs + 25 mm
12
8
-
4 :
01
’ 30
1
-20 Before (mln)
I
IO
I
0 Phase III onSet
I
+10
I ?? 20
After (min)
1
+30
01 -30’
I
-20 Before (min)
I -10
I Phaie Ill onset
1 *lo
1 ?? 20
I ?? 30
After (mini
Figure 3. Upper left panel: gastric acid output before and after onset of phase III activity (*p -C 0.05 vs. +l5 and +25 min, sign test). Upper right panel: duodenal bile acid output before and after onset of phase III activity (*p < 0.05 vs. +5, +15, and +25 min, sign test). Bottom right panel: duodenal trypsin output before and after the onset of phase III activity (*p < 0.05 vs. +15 and +25 min. ““p < 0.05 vs. +25 min, sign test). Bottom left panel: duodenal bicarbonate output before and after the onset of phase III activity (*p < 0.05 vs. -25, -15, and +25 min, sign test).
730
GASTROENTEROLOGY
KEANE ET AL.
Vol. 81,No.4
product of duodenal bile acid output and percentage of duodenogastric marker reflux. However, it should be noted that reflux of [14C]PEG into the stomach, 30 like that of bile acids or trypsin, is greater before T phase III. Yet, [l’C]PEG is constantly perfused into the duodenum and, therefore, not subjected to the 20 cyclic variations observed in bile acid or trypsin se10 cretion. This supports the concept that mechanical factors such as motor activity and/or the volume of duodenal contents are important determinants of Before Phase III After (min) onset (min) fasting refIux and that duodenogastric reflux is not Figure 4. Percentage duodenogastric reflux of [%]PEG before simply a reflection of secretory activity. and after the onset of phase III motor activity, ?? p < 0.05 The role of phase III activity as a clearing mechavs. +15 and +25 min (sign test). nism for refluxed duodenal contents from the stomach fits well with established concepts. Shortly after the description of interdigestive motor complexes by mined by the product of percent marker reflux and (17)and Grivel and Ruckebusch (16), Szurszewski duodenal bile acid output). The coefficient of correCode and Schlegel (19), by fluoroscopic observalations between observed and predicted reflux for tions, proposed a propulsive, cleansing function of the 4 individuals were 0.97, 0.64, 0.63, and 0.62. phase III, coining the term “intestinal housekeeper.” Subsequently, Mroz and Kelly (20)have shown in Discussion dogs that phase III of the interdigestive motor complex clears the stomach of large, inert solids which Our study shows a clear relationship among may have remained as a residue after ingestion of a gastrointestinal cyclic motor and secretory activity mixed meal containing nondigestible spheres. It apand duodenogastric reflux. A marked increase in pears from our observations that, in humans, gastric pancreatic enzyme and biliary secretions occurs phase III also plays a clearing role in evacuating during phase II (irregular activity) of the interfrom the stomach material refluxed during the predigestive motor complex. This parallel increase in ceding motor-secretory phase. motor and secretory activity apparently creates conThe experimental design of our study included ditions favoring duodenbgastric reflux which also three different approaches to evaluate the potential peaks during phase II. In contrast, phase III motor role of duodenal intubation and perfusion on duoactivity coincides with a reduction of reflux and a denogastric reflux and motor-secretory cycles. It apclearance from the stomach of duodenal secretions pears that a small transpyloric tube does not have a which refluxed during the preceding phase. significant effect on the pattern or magnitude of duoFrom animal studies it has been recognized that denogastric reflux. Further, duodenal perfusion rates duodenogastric reflux depends far more on the relaup to 3 ml/min had no appreciable effect on motor tive contractile pattern on either side of the pylorus than on the diameter of the pyloric ring itself (15,16). activity. Others, however, have shown that perfusion rates, faster than those used in the present Clearing of refluxed material associated with antral study, are likely to increase reflux of the perfused contractile activity has also been noted in the dog duodenal solution into the stomach (21). Earlier we (16). However, the relationship between cyclic motor-secretory activity and duodenogastric reflux has Bile acld cone not been previously examined. Our study suggests (pmol/ml) that fasting duodenogastric reflux in humans is the 14 result of a dual process: an increase in pancreatic 12 and biliary secretion into the duodenum which 10 makes these substances available for reflux and an 8 irregular motor activity (phase II) which creates the 6 necessary pressure gradients for reflux across the 4 antroduodenal junction to take place. 2 I I I 1 I I / 1 The motor and secretory components of reflux 30 20 -10 0 +10 +20 *a0 are, to a large extent, dissociable but both contribute PhaSe III Before After onset (min) Wnl to reflux. We have obtained evidence for this suppoFigure 5. Gastric bile acid concentration before and after onset sition because gastric bile acid content before and of phase III activity during design 1 and 2 studies; tp < after the onset of phase III (measured by design 3) 0.05vs.+15,+25 min; *p < 0.05vs.+15 min (paired tcould be accurately predicted by determining the test). percent I’C-peg refluxed 40 r
T /A 1
T
11
October 1981
INTERDIGESTIVE
have shown that negative aspiration and duodenal perfusion do not alter basal intraluminal pressures in the antrum and duodenum (11). Our studies open a path for evaluating the relationship of motor-secretory cycles to duodenogastric reflux in pathophysiokogic conditions. Gastric ulcer and esophagitis are among the diseases in which gastric emptying abnormalities and increased duodenogastric reflux have been described (2); they deserve further investigation in light of the findings reported here regarding relationships between reflux and the interdigestive migrating motor complex. Nevertheless, establishing abnormalities in these areas may be difficult. Even apparently healthy individuals exhibit great variability in their patterns of motor-secretory activity and duodenogastric reflux. Occasionally, individuals such as 1 subject in our study are found who have rare or poorly developed interdigestive motor complexes. In those instances, secretory cycles also tend to be less distinct and relationships between motor-secretory events and duodenogastric reflux cannot be easily established. Much more work is needed to unequivocally determine the limits between normal and abnormal physiology in the interdigestive period.
References 1. Wormsley
RG. Aspects of duodenogastric reflux in man. Gut 1972;13:243-50. 2. Rees WDW, Rhodes J. Bile reflux in gastro-oesophageal disease. Clin Gastroenterol 1977;6:179-200. 3. Grant R, Grossman MI, Wang KJ, et al. The cytolic action of some gastrointestinal secretions and enzymes on epithelial cells of the gastric and duodenal mimosa. J Cell Comp Physiol 1951;37:137-61. 4. Davenport HN. Destruction of the gastric mucosal barrier by detergents and urea. Gastroenterology 1968;34:175-81. 5. Lawson HH. Effect of duodenal contents on the gastric mucosa under experimental conditions. Lancet 1964;1:469-72. 6. Delaney JP, Cheng JWB, Butler BA, et al. Gastric ulcer and regurgitation gastritis. Gut 1970;11:715-9.
ACTIVITY AND DLJODENOGASTRIC
REFLUX
731
7. Kaye MD, Mahter SJ, Showalter JP. Manometric studies on the human pylorus. Gastroenterology 1976:70:477-80. 8. DiMagno EP, Hendricks JC, Go VLW, et al. Relationships among canine fasting pancreatic and biliary secretions, pancreatic duct pressure and duodenal phase III motor activity--Boldyreff revisited. Dig Dis Sci 1979;24:689-93. 9. Keane FB, DiMagno EP, Dozois RR, et al. Relationships among canine interdigestive exocrine pancreatic and biliary flow, duodenal motor activity, plasma pancreatic polypeptide, and motilin. Gastroenterology 1980;78:310-6. 10. Vantrappen GR, Peeters TL, Janssens J. The secretory component of the interdigestive migrating motor complex in man. Stand J Gastroenterol 1979;14:663-7. 11. Rees WDW, Go VLW, Malagelada J-R. Simultaneous measurement of antroduodenal motility, gastric emptying, and duodenogastric reflux in man. Gut 1979;20:963-70. 12. Arndorfer RC, Stef JJ, Dodds WJ, et al. Improved infusion system for intraluminal esophageal manometry. Gastroenterology 1977;73:23-7. 13. Talalay P. Enzymic analysis of steroid hormones. Methods Biochem Anal 1960;8:119-43. 14. Pelot D, Grossman MI. Distribution and fate of pancreatic enzymes in small intestine of the rat. Am J Physiol1962;202:2858. G, et al. Duodeno15. Sonnenberg A, Lepsien G, Schattenmann gastric reflux in the dog following pharmacological antral and pyloric inhibition (abstr). Proceedings of the 7th International Symposium on Gastrointestinal Motility, Iowa City, Iowa, 1979:38. 16. Munk JF, Johnson AG. Effects of duodenal and antral pacing on pyloric reflux in the cat (abstr). Proceedings of the 7th International Symposium on Gastrointestinal Motility, Iowa City, Iowa, 1979:40. 17. Szurszewski JH. A migrating electric complex of the canine small intestine. Am J Physiol 1969;217:1757-63. 18. Grivel ML, Ruckebusch Y. The propagation of segmental contractions along the small intestine. J Physiol (Lond) 1972; 227:611-25. interdigestive 19. Code CF, Schlegel JF. The gastrointestinal housekeeper: motor correlates of the interdigestive myoelectric complex of the dog. In: Daniel EE, ed. Proceedings of the Fourth International Symposium on Gastrointestinal Motility. Vancouver, British Columbia: Mitchell Press, 1973:6313. 20. Mroz CT, Kelly KA. The role of the extrinsic antral nerves in the regulation of gastric emptying. Surgery 1977;145:369-77. 21. Gustke RF, Varma RR, Soergel KH. Gastric reflux during perfusion of the proximal small bowel. Gastroenterology 1970;59:894-5.
1981;81:728-31
Duodenogastric Reflux in Humans: Its Relationship to Fasting Antroduodenal Motility and Gastric, Pancreatic, and Biliary Secretion FRANK B. KEANE, EUGENE and JUAN-R. MALAGELADA Gastroenterology
Unit,
Mayo
Clinic
and
P. DIMAGNO, Mayo
Duodenogastric reflux may have pathophysiologic importance, but its mechanism is poorly understood. We propose that duodenogastric reflux and periodic changes in motor and secretory activity of the upper gut during fasting may be related. Therefore we determined the relationships between duodenogastric reflux and interdigestive motor-secretory cycles in a group of 6 healthy individuals, on each of whom we performed three 7-h studies in random order on separate days. In all studies gastric intubation and antral pressure recordings were performed. However, in design 1 we used a slow duodenal perfusion rate of rC]PEG in saline (0.25 mI/min) while in design 3 we used a fast perfusion rate (3.0 ml/min). Duodenal pressures were also recorded during these designs. In design 2 no transpyloric tubes were present. Our study shows that, in humans, fasting duodenogastric reflux of bile and pancreatic juice is cyclic and closely related to the interdigestive migrating motor complex. Reflux is highest during late phase II (when secretory activity is also on the rise) and lowest after phase III. One of the important functions of the migrating motor complex in humans may be to clear the stomach of refluxed duodenal secretions. In humans, reflux of duodenal contents into the stomach occurs sporadically during fasting and after meals (1). Increased duodenogastric reflux has been found in patients with gastric ulcer, esophagitis, and gallstone dyspepsia (2). Bile acids and pancreatic enzymes, normally found in the duodenum, have been Received August 8, 1980. Accepted May 22, 1981. Address requests for reprints to: Juan-R. Malagelada, M.D., Gastroenterology Unit, Mayo Clinic, Rochester, Minnesota 55905. This investigation was supported in part by Grants CA 25064 and AM 26428 from the National Institutes of Health, Public Health Service, Bethesda, Maryland. Dr. Malagelada is the recipient of Research Career Development Award AM 00330 from the National Institutes of Health. 0 1981 by the American Gastroenterological Association OOlS-5085/81/100726-08$02.50
Foundation,
Rochester,
Minnesota
shown experimentally to be capable of damaging the gastric and esophageal mucosa (3-6). However, while the mechanism of duodenogastric reflux may be both physiologically and pathophysiologically important, it is poorly understood (7). Two premises, however, seem valid: first, duodenal contractile activity is the propulsive force for duodenal contents to move in a retrograde direction, and, second, the pylorus must be open, at least partially, for this material to enter the stomach. During fasting, gastrointestinal motility is organized in alternating cycles of activity and quiescence. Characteristic features of the former are its distal propagation from the stomach to the ileum and its association with an increase in gastric, pancreatic, and biliary secretion (8-10). These current concepts of fasting motor and secretory activity of the gut suggest that duodenogastric reflux is also related to interdigestive motor-secretory cycles. Preliminary data from our laboratory have indicated that phase III (intense, propagated activity) of the interdigestive motor cycle may play an antireflux role (11). However, very little work along these lines has been performed to date. In the present study we have attempted to characterize the temporal relationships between motor-secretory cycles and duodenogastric reflux in health. In order to resolve potential technical difficulties and theoretical objections that might arise secondary to the use of gastroduodenal intubation and perfusion, we employed three different experimental designs to quantify motor and secretory activity and duodenogastric reflux in each individual.
Materials, Design
Methods,
and Experimental
Studies were performed on 6 healthy aged 23-53 yr. Informed consent was obtained.
volunteers In random
October 1981
INTERDIGESTIVE
order, three studies were performed on each volunteer on 3 different days. The three studies involved different assemblies of gastrointestinal tubes (Figure 1). After an overnight fast a volunteer was asked to swallow one of these assemblies. After the position of the tubes was confirmed fluoroscopically, the volunteer assumed a comfortable semirecumbent position. The purpose of using three designs in this study was to: (a) evaluate simultaneous gastric and duodenal motor-secretory activity under near physiologic conditions; (b) determine whether a transpyloric tube alters duodenogastric reflux; and (c) determine whether a marked increase in duodenal perfusion rate would change our results. Accordingly, in design 1 very low gastric and duodenal perfusion rates were used, in design 2 we omitted the duodenal tube, and in design 3 duodenal perfusion rates were 12-fold greater (3 ml/min) than in design 1.
Design 1 Study (Slow Duodenal
Perfusion)
Two sump-type triple-lumen polyvinyl tubes were positioned so that the tip of one lay at the ligament of Treitz and the tip of the other was in the dependent part of the stomach (Figure 1). The duodenal tube contained two miniature strain gauge pressure transducers (Millar Instruments, Houston, Tex., Model PC 350), one positioned in the antrum and the other at the ligament of Treitz. One lumen of the duodenal tube was located opposite the ampulla of Vater and was used to perfuse polyethylene-l-2[‘4C]glycol ([Y]PEG) (New England Nuclear, Boston, Mass.), 0.2 &i/ml 154 mM NaCl, at 0.25 ml/min. Perfusion was carried out with a low compliance perfusion system (12) which permitted simultaneous pressure monitoring. A strain gauge (Statham Instruments, Hato Rey, Puerto Rico) attached to a Gould recorder was used to record pressures. The second lumen of the duodenal tube was used to continuously aspirate fluid at the ligament of Treitz. The third lumen acted as an air vent. A fingercot containing 0.75 ml of mercury was attached to the distal end of the tube to facilitate its passage into the duodenum. The gastric tube was also used for perfusion and aspiration. The perfusate contained a second nonabsorbable marker, polyethylene-l-2-[3H]-glycol ([3H]PEG) (New Engxstm
1-3
PERFUSION
RATE
TRANSPYLORIC NO TRANSPYLORIC NO DUODENAL
REFLUX
727
land Nuclear, Boston, Mass.), 0.2 &i/ml 154 mM NaCl. This solution was perfused by means of a syringe pump (Harvard Apparatus, Dover, Mass.) through a port 10 cm proximal to the tip of the gastric tube, at a rate of 0.072 ml/ min. Meanwhile, gastric contents were continuously aspirated from the tip of the gastric tube. Gastric and duodenal aspirates were simultaneously collected every 10 min. Gastric juice was analyzed for pH, titratable acidity, bile acids, [3H]PEG, and [Y]PEG. Duodenal samples were analyzed for pH, bile acids, trypsin, bicarbonate, and [‘*C]PEG.
Design 2 Study (No Duodenal
Tubes)
Two polyvinyl tubes were placed in the stomach. To one of these gastric tubes two miniature strain gauge pressure transducers were attached, 3.5 cm apart, and positioned so as to record antral motility (Figure 1). The second gastric tube was a triple-lumen sump tube. As in the previous study design, [3H]PEG was perfused at 0.072 ml/ min, 10 cm proximal to the tip of this tube. From the tip, gastric juice was sampled at IO-min intervals and analyzed for pH, titratable acidity, bile acids, and [3H]PEG.
Design 3 Study (Fast Duodenal
Perfusion)
A duodenal tube and a gastric tube similar to those used in the design 1 study were used. However, the gastric tube was used only for sampling gastric contents and no marker was perfused into the stomach. The duodenal tube carried two miniature strain gauge pressure transducers positioned 22 cm apart such that one lay in the antrum and the other lay in the second part of the duodenum opposite the ampulla of Vater (Figure 1). Polyethylene glycol (PEG 4000), 5 g/l 154 mM NaCl, heated to 37% was perfused at 3 ml/min through a polyvinyl tube whose opening also lay opposite the ampulla of Vater. Through another fine polyvinyl tube attached to the duodenal tube, also opening at the second portion of the duodenum, 0.5-ml pulses of a second nonabsorbable marker, [3H]PEG (3 pCi/ ml), were injected at intervals and flushed with 1.5 ml saline, a volume equal to the dead space volume of the tube. The interval between injection and subsequent recovery
e-
-JIG-
TRANSPYImORIC TlJBE t.OW DUODENAi
ACTIVITY AND DUODENOGASTRIC
TUBE
PERFUSION
TUBE
FAST DUODENAL PERFUSICN
RATF
Figure 1. Assemblies used to simultaneously measure and validate measurements of gastric, biliary, and pancreatic secretion and antroduodenal motility.
728
GASTROENTEROLOGY
KEANE ET AL.
of this marker in duodenal collections taken every 5 min for 20 min after injection represented the time necessary for duodenal contents to travel from the ampulla of Vater through the duodenum and collecting catheter. Duodenal aspirates were collected continuously at the ligament of Treitz every 10 min and analyzed for pH, bile acids, trypsin, bicarbonate, PEG 4000, and [3H]PEG. Gastric samples taken during each lo-min interval were analyzed for pH, titratable acidity, and bile acids. All of the studies were begun after a 1-h equilibration period and continued for approximately 7 h (average duration for design 1, 415 f 14 min, SEM, n = 6; for design 2, 440 f 15 min; and for design 3, 390 f 11 min). Each Millar transducer was connected to a Gould four-channel recorder (Gould Model 2400, Cleveland, Ohio) via a control unit (Millar Instruments, Houston, Tex., Model TC 100) that provided zero and 100 mmHg calibration at room temperature (thermal stability of sensors = 1 mmHg over room temperature, range 23’-38’C). The lo-min sampling intervals were marked on each motility record to facilitate analysis.
Analytic
Procedures
Using a glass electrode, pH was measured. Titratable acidity of the samples was determined by titration with NaOH. Bile acid concentrations were determined by the steroid dehydrogenase method (13) trypsin by titrimetric assay (14) and bicarbonate by CO, analyzer (Ericsen Instruments, Ossining, N.Y., Model E-109). [l’C]PEG and [3H]PEG were assessed by /3-scintillation counting while PEG 4000 was determined turbidimetrically. Outputs were calculated based on the recovery of the nonabsorbable markers. During design 1, the slow duodenal perfusion study, sample amounts obtained from the ligament of Treitz were frequently insufficient for analytical purposes (70% of the lo-min samples). Antroduodenal motility records were divided into lomin periods corresponding to the sampling intervals. Records were analyzed for the start of phase III motor activity in the second part of the duodenum, phase III being defined as regular duodenal contractions at a frequency not less than 8/min and continuing for at least 3 min. During design 1 and 3 studies, propagation of the activity front (phase III) could be followed from the antrum to the duodenum. In these studies we observed that the end of phase III activity in the antrum corresponded to the start of phase III activity in the second part of the duodenum. Therefore, during design 2 studies, when recordings were only obtained from the antrum, the onset of phase III activity in the second part of the duodenum was defined as the end of phase III in the antrum. The end of phase III in the antrum was identified as the cessation of 3 contractions/min.
Statistical
Methods
For the purpose of analysis, collection periods were grouped in relation to appearance of phase III activity in the second portion of the duodenum. Three lo-min collection periods before and two lo-min periods after the
Vol. Al. No. 4
onset of phase III activity were considered to be associated with the onset of phase III activity while the remain-
ing lo-min collections were considered to be unassociated. Results were averaged across episodes of phase III activity for the 6 patients and this mean value per patient was used in subsequent statistical analysis. Significant differences among individual values for marker reflux, bile acid output, and other secretory parameters during the three IO-min intervals before and after the onset of phase III were determined by using the sign test.
Results Motor Activity During most experiments phase III activity recurred at intervals which were not always regular. One volunteer had less frequent activity fronts than the others-two on one day, one on the other, and none on the third day. During design 1 and design 3 studies, over 90% of the activity fronts were propagated from the antrum to the duodenum. The mean frequency of phase III activity for each volunteer during each experiment was just over 3 with a range varying from 0 to 5. The frequency of phase III activities for each of the experimental designs was very similar (average number of phase III activities was 3.3 + 0.6, SEM, n = 6 for design 1, 3.2 + 0.4 for design 2, and 3.3 + 0.7 for design 3) suggesting that the frequency of activity fronts was not altered by a transpyloric tube or by slow (0.25 ml/ min) or fast (3 ml/min) duodenal perfusion. Secretion Increases in gastric acid output (from design 1 and design 2 studies), and duodenal bile acid, trypsin, and bicarbonate outputs (from design 3 study) were associated with the onset of phase III motor activity (Figure 2). Mean gastric acid output and duodenal bicarbonate outputs were maximal during the IO-min period after the onset of phase III while mean duodenal bile acids and trypsin outputs were higher prior to the onset of phase III activity (Figure 3). Average values for each of the variables 30 min before and after phase III differed significantly (p < 0.05) by the sign test. During the design 3 study the transit time from the ampulla of Vater to the collection site for duodenal samples was measured using injected pulses of [3H]PEG. The phase of motor activity at the time of injection was defined as follows: phase I was a flat tracing reflecting absent duodenal motor activity, phase II reflected irregular contractions, and phase III reflected a burst of regular contractions as previously described. Peak recovery of marker pulses at the ligament of Treitz occurred within 10 min re-
INTERDIGESTIVE ACTIVITY AND DUODENOGASTRlC
1981
October
REFLUX
729
Representative example showing phases of motor activity in association with gastric acid output and duodenal outputs of trypsin and bile acids. Regular recurring fronts. Phase I, white portion; phase II, stippled portion; phase III, solid black portion.
L
I
I
I
1
2
1
-3e--
3--'
Time (hours)
gardless of the motor phase and no statistically significant differences were noted (data not shown). By 30 min after injection of marker pulses a mean of 58% had been recovered after injection during phase I. 59.2% during phase II, and 57.2% during phase III. Duodenogastric
Reflux
Duodenogastric reflux of [“CIPEG was meaduring the design 1 study when the stomach sured was continuously aspirated. Reflux was maximal prior to the onset of the activity front. Following the onset of phase 111activity intragastric [W]PEG was cleared from the stomach (Figure 4). Average
Bile acid out ut (pmol/min P
Gastrcc acid output (mEq/mm) 4 $ z -k *SE * = p
01 HCOj
,
’
[Y]PEG reflux 30 min before and after phase III differed significantly by the sign test (p c 0.05). Mean intragastric bile acid concentration, measured for each volunteer during the three studies, was maximal prior to the onset of phase III activity (Figure 5). There were no consistent differences in mean gastric bile acid concentrations among the three study designs when the means from each design were compared (data not shown). In 4 individuals who had three or more episodes of phase III during day 3, we compared actual intragastric bile acid content (determined from gastric samples by multiplying intragastric bile concentration by gastric volume) with the predicted bile acid reflux (deter-
05
15and+25min
I
1
I
I
I
0
,_.-__L
VS + 5, + 15 + 25 min
...._..-L__i
output
vD”“mi”)
I:
F
T
16
tSE 05vs+
T **=p< + =p’
15and+25mm
05 vs + 25 mm
12
8
-
4 :
01
’ 30
1
-20 Before (mln)
I
IO
I
0 Phase III onSet
I
+10
I ?? 20
After (min)
1
+30
01 -30’
I
-20 Before (min)
I -10
I Phaie Ill onset
1 *lo
1 ?? 20
I ?? 30
After (mini
Figure 3. Upper left panel: gastric acid output before and after onset of phase III activity (*p -C 0.05 vs. +l5 and +25 min, sign test). Upper right panel: duodenal bile acid output before and after onset of phase III activity (*p < 0.05 vs. +5, +15, and +25 min, sign test). Bottom right panel: duodenal trypsin output before and after the onset of phase III activity (*p < 0.05 vs. +15 and +25 min. ““p < 0.05 vs. +25 min, sign test). Bottom left panel: duodenal bicarbonate output before and after the onset of phase III activity (*p < 0.05 vs. -25, -15, and +25 min, sign test).
730
GASTROENTEROLOGY
KEANE ET AL.
Vol. 81,No.4
product of duodenal bile acid output and percentage of duodenogastric marker reflux. However, it should be noted that reflux of [14C]PEG into the stomach, 30 like that of bile acids or trypsin, is greater before T phase III. Yet, [l’C]PEG is constantly perfused into the duodenum and, therefore, not subjected to the 20 cyclic variations observed in bile acid or trypsin se10 cretion. This supports the concept that mechanical factors such as motor activity and/or the volume of duodenal contents are important determinants of Before Phase III After (min) onset (min) fasting refIux and that duodenogastric reflux is not Figure 4. Percentage duodenogastric reflux of [%]PEG before simply a reflection of secretory activity. and after the onset of phase III motor activity, ?? p < 0.05 The role of phase III activity as a clearing mechavs. +15 and +25 min (sign test). nism for refluxed duodenal contents from the stomach fits well with established concepts. Shortly after the description of interdigestive motor complexes by mined by the product of percent marker reflux and (17)and Grivel and Ruckebusch (16), Szurszewski duodenal bile acid output). The coefficient of correCode and Schlegel (19), by fluoroscopic observalations between observed and predicted reflux for tions, proposed a propulsive, cleansing function of the 4 individuals were 0.97, 0.64, 0.63, and 0.62. phase III, coining the term “intestinal housekeeper.” Subsequently, Mroz and Kelly (20)have shown in Discussion dogs that phase III of the interdigestive motor complex clears the stomach of large, inert solids which Our study shows a clear relationship among may have remained as a residue after ingestion of a gastrointestinal cyclic motor and secretory activity mixed meal containing nondigestible spheres. It apand duodenogastric reflux. A marked increase in pears from our observations that, in humans, gastric pancreatic enzyme and biliary secretions occurs phase III also plays a clearing role in evacuating during phase II (irregular activity) of the interfrom the stomach material refluxed during the predigestive motor complex. This parallel increase in ceding motor-secretory phase. motor and secretory activity apparently creates conThe experimental design of our study included ditions favoring duodenbgastric reflux which also three different approaches to evaluate the potential peaks during phase II. In contrast, phase III motor role of duodenal intubation and perfusion on duoactivity coincides with a reduction of reflux and a denogastric reflux and motor-secretory cycles. It apclearance from the stomach of duodenal secretions pears that a small transpyloric tube does not have a which refluxed during the preceding phase. significant effect on the pattern or magnitude of duoFrom animal studies it has been recognized that denogastric reflux. Further, duodenal perfusion rates duodenogastric reflux depends far more on the relaup to 3 ml/min had no appreciable effect on motor tive contractile pattern on either side of the pylorus than on the diameter of the pyloric ring itself (15,16). activity. Others, however, have shown that perfusion rates, faster than those used in the present Clearing of refluxed material associated with antral study, are likely to increase reflux of the perfused contractile activity has also been noted in the dog duodenal solution into the stomach (21). Earlier we (16). However, the relationship between cyclic motor-secretory activity and duodenogastric reflux has Bile acld cone not been previously examined. Our study suggests (pmol/ml) that fasting duodenogastric reflux in humans is the 14 result of a dual process: an increase in pancreatic 12 and biliary secretion into the duodenum which 10 makes these substances available for reflux and an 8 irregular motor activity (phase II) which creates the 6 necessary pressure gradients for reflux across the 4 antroduodenal junction to take place. 2 I I I 1 I I / 1 The motor and secretory components of reflux 30 20 -10 0 +10 +20 *a0 are, to a large extent, dissociable but both contribute PhaSe III Before After onset (min) Wnl to reflux. We have obtained evidence for this suppoFigure 5. Gastric bile acid concentration before and after onset sition because gastric bile acid content before and of phase III activity during design 1 and 2 studies; tp < after the onset of phase III (measured by design 3) 0.05vs.+15,+25 min; *p < 0.05vs.+15 min (paired tcould be accurately predicted by determining the test). percent I’C-peg refluxed 40 r
T /A 1
T
11
October 1981
INTERDIGESTIVE
have shown that negative aspiration and duodenal perfusion do not alter basal intraluminal pressures in the antrum and duodenum (11). Our studies open a path for evaluating the relationship of motor-secretory cycles to duodenogastric reflux in pathophysiokogic conditions. Gastric ulcer and esophagitis are among the diseases in which gastric emptying abnormalities and increased duodenogastric reflux have been described (2); they deserve further investigation in light of the findings reported here regarding relationships between reflux and the interdigestive migrating motor complex. Nevertheless, establishing abnormalities in these areas may be difficult. Even apparently healthy individuals exhibit great variability in their patterns of motor-secretory activity and duodenogastric reflux. Occasionally, individuals such as 1 subject in our study are found who have rare or poorly developed interdigestive motor complexes. In those instances, secretory cycles also tend to be less distinct and relationships between motor-secretory events and duodenogastric reflux cannot be easily established. Much more work is needed to unequivocally determine the limits between normal and abnormal physiology in the interdigestive period.
References 1. Wormsley
RG. Aspects of duodenogastric reflux in man. Gut 1972;13:243-50. 2. Rees WDW, Rhodes J. Bile reflux in gastro-oesophageal disease. Clin Gastroenterol 1977;6:179-200. 3. Grant R, Grossman MI, Wang KJ, et al. The cytolic action of some gastrointestinal secretions and enzymes on epithelial cells of the gastric and duodenal mimosa. J Cell Comp Physiol 1951;37:137-61. 4. Davenport HN. Destruction of the gastric mucosal barrier by detergents and urea. Gastroenterology 1968;34:175-81. 5. Lawson HH. Effect of duodenal contents on the gastric mucosa under experimental conditions. Lancet 1964;1:469-72. 6. Delaney JP, Cheng JWB, Butler BA, et al. Gastric ulcer and regurgitation gastritis. Gut 1970;11:715-9.
ACTIVITY AND DLJODENOGASTRIC
REFLUX
731
7. Kaye MD, Mahter SJ, Showalter JP. Manometric studies on the human pylorus. Gastroenterology 1976:70:477-80. 8. DiMagno EP, Hendricks JC, Go VLW, et al. Relationships among canine fasting pancreatic and biliary secretions, pancreatic duct pressure and duodenal phase III motor activity--Boldyreff revisited. Dig Dis Sci 1979;24:689-93. 9. Keane FB, DiMagno EP, Dozois RR, et al. Relationships among canine interdigestive exocrine pancreatic and biliary flow, duodenal motor activity, plasma pancreatic polypeptide, and motilin. Gastroenterology 1980;78:310-6. 10. Vantrappen GR, Peeters TL, Janssens J. The secretory component of the interdigestive migrating motor complex in man. Stand J Gastroenterol 1979;14:663-7. 11. Rees WDW, Go VLW, Malagelada J-R. Simultaneous measurement of antroduodenal motility, gastric emptying, and duodenogastric reflux in man. Gut 1979;20:963-70. 12. Arndorfer RC, Stef JJ, Dodds WJ, et al. Improved infusion system for intraluminal esophageal manometry. Gastroenterology 1977;73:23-7. 13. Talalay P. Enzymic analysis of steroid hormones. Methods Biochem Anal 1960;8:119-43. 14. Pelot D, Grossman MI. Distribution and fate of pancreatic enzymes in small intestine of the rat. Am J Physiol1962;202:2858. G, et al. Duodeno15. Sonnenberg A, Lepsien G, Schattenmann gastric reflux in the dog following pharmacological antral and pyloric inhibition (abstr). Proceedings of the 7th International Symposium on Gastrointestinal Motility, Iowa City, Iowa, 1979:38. 16. Munk JF, Johnson AG. Effects of duodenal and antral pacing on pyloric reflux in the cat (abstr). Proceedings of the 7th International Symposium on Gastrointestinal Motility, Iowa City, Iowa, 1979:40. 17. Szurszewski JH. A migrating electric complex of the canine small intestine. Am J Physiol 1969;217:1757-63. 18. Grivel ML, Ruckebusch Y. The propagation of segmental contractions along the small intestine. J Physiol (Lond) 1972; 227:611-25. interdigestive 19. Code CF, Schlegel JF. The gastrointestinal housekeeper: motor correlates of the interdigestive myoelectric complex of the dog. In: Daniel EE, ed. Proceedings of the Fourth International Symposium on Gastrointestinal Motility. Vancouver, British Columbia: Mitchell Press, 1973:6313. 20. Mroz CT, Kelly KA. The role of the extrinsic antral nerves in the regulation of gastric emptying. Surgery 1977;145:369-77. 21. Gustke RF, Varma RR, Soergel KH. Gastric reflux during perfusion of the proximal small bowel. Gastroenterology 1970;59:894-5.