DUODENAL-GASTRIC REFLUX AND SLOWED GASTRIC EMPTYING BY ELECTRICAL PACING OF THE CANINE DUODENAL PACESETTER POTENTIAL

Transcription

1 GASTROENTEROLOGY 72: , 1977 Copyright 1977 by The Williams & Wilkins Co. Vol. 72,No. 3 Printed in U.S.A. DUODENAL-GASTRIC REFLUX AND SLOWED GASTRIC EMPTYING BY ELECTRICAL PACING OF THE CANINE DUODENAL PACESETTER POTENTIAL KEITH A. KELLY, M.D., AND CHARLES F. CODE, M.D., PH.D. Section of Gastrointestinal Surgical Research and Department of Physiology and Biophysics, Mayo Clinic and Mayo Foundation, Rochester, Minnesota In 6 (10 to 12 kg) mongrel female dogs, silver electrodes for recording electrical activity and for pacing of pacesetter potentials (PP) were implanted on the stomach and duodenum and a catheter for intraluminal instillations was inserted into the duodenum. Beginning 2 weeks after operation, electrical recordings were made intermittently from the fasted, conscious dogs with no pacing and during pacing of the PP in the proximal and in the distal duodenum. A suspension of BaS04 injected into the unpaced duodenum and observed cinefluoroscopically was swept quickly from the duodenum into the jejunum; little or none passed orad into the proximal duodenum, and BaS04 rarely entered the stomach. Only 1 to 3% of a duodenal infusate of 154 mm NaCl with [14C] polyethylene glycol (2 ml per min) appeared in the stomach after 15 min. The results during proximal duodenal pacing were the same as with no pacing. However, distal duodenal pacing, which reversed the direction of propagation of the duodenal PP's, caused duodenal-gastric reflux of BaS04 in every dog and forced about 30% of the duodenal infusate into the stomach during fasting and during gastric emptying of 400 ml of 154 mm NaCl; at the same time, the rate of emptying of the gastric instillate was slowed about 25%. The pacesetter potential (electrical control activity, slow wave) ofthe duodenum can be controlled by electrical pacing. We 3 found, as did Sarna and Daniel,4 that electrical pacing of the proximal duodenum increased the frequency of the duodenal pacesetter potential (PP) and did not change its aborad direction of propagation. In contrast, electrical pacing of the distal duodenum not only increased the frequency of the duodenal PP but also reversed its direction of propagation. The PP was propagated orally from the distal site of pacing to the pylorus. The question asked in the present experiments was: "Would distal duodenal pacing, with orad propagation of the duodenal PP, result in duodenal-gastric reflux and slowed gastric empyting?" Methods Construction of electrodes. Bipolar electrodes were used for pacing and unipolar electrodes for concurrent recording of Received May 7, Accepted September 28, Abstracts of portions of this work have been published Address requests for reprints to: Dr. K. A. Kelly, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 5590l. This investigation was supported in part by Research Grants AM and AM from the National Institutes of Health, United States Public Health Service, and by the Mayo Foundation. Dr. Code's present address is: Center for Ulcer Research and Education (CURE), Building 115, Veterans Administration Wadsworth Hospital Center, Los Angeles, California The authors gratefully acknowledge the help of J. F. Schlegel and J. H. Steinbach during the cinefluorographic tests. 429 electrical activity. The electrodes, which were similar to those previously used in our laboratory,3 were made of silver wire (diameter 1 mm) that projected 1 mm from an acrylic disc. The shafts of the electrodes were insulated with vinyl paint, but their tips were exposed and were coated electrolytically with silver chloride. The shafts of bipolar electrodes were spaced 5 mm apart in the acrylic disc. The electrodes were soldered to a Teflon-insulated copper wire, and the junction was encased within the acrylic discs. The copper wires connected the discs to a pin of a multipinned socket that was mounted within a stainless steel cannula. Placement of electrodes. Six healthy mongrel female dogs (10 to 12 kg) were used. Each dog was anesthetized with sodium pentobarbital while the electrodes were sewn to the serosal surface of the stomach and duodenum with 4-0 Dacron interrupted sutures, using sterile technique. The bipolar electrodes were oriented along the longitudinal axis of the bowel. Three monopolar gastric electrodes were applied 6,4, and 2 cm proximal to the gastroduodenal junction on the anterior serosal surface of the stomach midway between the greater and lesser curvatures (fig. 1). One bipolar pacing electrode was implanted 1 cm distal to the gastroduodenal junction on the anterior wall ofthe duodenum, and a second was positioned 26 cm more distal. Four monopolar electrodes (5 cm apart) were then spaced in between the two bipolar electrodes, and a fifth monopolar electrode was placed 5 cm distal to the caudad bipolar electrode. The stainless steel cannula carrying the leads to the surface was positioned in the right anterior abdominal wall and secured with wire sutures. Placement of duodenal catheter. A polyvinyl, flanged catheter (outside diameter 2 mm) was inserted into the lumen of the midduodenum, 13.5 cm distal to the gastroduodenal junction,

2 430 KELLY AND CODE Vol. 72, No.3 I POSITION OF ELECTRODES AND DUODENAL CATHETER Catheter ELECTRODES 'i'= Recording 1t= Stimulating FIG. 1. Positions of gastric and duodenal recording and stimulating electrodes and of duodenal catheter. at the same operation (fig. 1). The perforated flange was sewn to the serosal surface of the duodenum with nonabsorbable sutures. The opposite end of the catheter was brought to the surface via a connector housed within a stainless steel cannula that was positioned in the left anterior abdominal wall and secured with sutures. Starting 2 weeks after the operation, studies were made intermittently for 3 to 6 months. Each dog was always fully conscious and had fasted for 18 to 24 hr. Cinefluorographic tests. The dogs were trained to lie quietly in the supine position on the cinefluorographic table. Electrical activity was recorded by connecting four duodenal and two gastric, unipolar electrodes to a six-channel, high impedance, alternating-current, pen-writing recorder (Brush, Mark 260, Brush Instruments, Cleveland, Ohio). The metallic cannula in the abdominal wall was used as the indifferent electrode. The amplifiers of the recorder had a time constant of 1 sec and were capable of recording frequencies up to 100 Hz. The fasting pattern of intestinal electrical activity was identified in each dog. The times when random action potentials, or phase II of the interdigestive myoelectrical complex, 5 were present in the proximal small intestine were noted. Fivemilliliter boluses of barium sulfate were then injected into the lumen via the duodenal catheter, and the pattern and direction of movement of the duodenal content were documented by cinefluorography. Rectangular pulses of direct current (strength, 2 rna; duration, 50 msec; frequency, 19 to 21 per min) were then applied to the proximal duodenal bipolar electrode. The parameters of pacing were chosen because they consistently entrained the duodenal PP in every dog. The pulses increased the frequency of the duodenal PP from a mean of 18 cycles per min before pacing to a mean of20 cycles per min with pacing (P < 0.05 in every dog, Student's t-test for paired data) but did not change its aborad direction of propagation (fig. 2). The mean frequency (5 cycles per min) and aborad direction of propagation of the gastric PP were also unchanged. Electrical isolation of the stimulator from the recording apparatus minimized the registration of stimulus artifact. The duodenal injection of barium sulfate with cinefluorography was repeated during proximal duodenal pacing. The test with barium was repeated a third time, but on this occasion, the distal duodenal electrodes were paced. Distal pacing not only increased the frequency of the duodenal PP from a mean of 18 cycles per min to a mean of 20 cycles per min (P < 0.05 in every dog) but also reversed the direction of propagation of the PP from aborad to orad in the duodenum proximal to the site of pacing (fig. 2). However, distal duodenal pacing did not alter the frequency or aborad direction of propagation of the gastric PP. The sequence of tests with no pacing, proximal duodenal pacing, and distal duodenal pacing was done on two occasions in dogs 1 through 4. Duodenal perfusion tests. These tests were performed to quantitate the effect of pacing on duodenal-gastric reflux when the stomach was empty. The dogs were positioned in Pavlov slings where they rested or stood comfortably, while gastroduodenal electrical recordings were made as described. Phase II of the interdigestive myoelectrical complex was again identified, at which time a duodenal infusion of 154 mm NaCI with % [1,2-14C]polyethylene glycol (New England Nuclear, Boston, Mass., mean molecular weight ) was started through the duodenal catheter at 2 ml per min and continued for 15 min, when an oral gastric tube was inserted and the gastric content was recovered. The stomach was washed with 100 ml of water and the wash was aspirated. After a IS-min period of rest, the test was repeated during proximal pacing of the duodenum. Mter another IS-min rest, the experiment was repeated a third time during distal pacing. The sequence of no pacing, proximal pacing, and distal pacing was randomized and was done on at least five occasions in each of 4 dogs (dogs 3 through 6). Gastric emptying tests. These tests were performed to quantitate the effect of pacing on duodenal-gastric reflux during gastric emptying. The duodenal perfusion experiments were repeated as described during phase I or II of the interdigestive myoelectrical complex, but at the onset of the infusions the E, 11? E2 16? E, 26 I S2,sec E2 Stimuli! E 2 1f [3_ W (17cycles/min) tacing at 51 (\9 cycles/m in)!.).. 1\ r r - L, - j \ ~ h / ~ L ~ J \ ~ - ~ ~. \ ~ - - > F - ~ ' ir Pacing at 52 (19cycles/min) -tt-f "'f['". FIG. 2. Recordings of duodenal electrical activity. Electrical pacing of both the proximal (8,) and the distal duodenum (8 2 ) increases the frequency of the duodenal pacesetter potential (PP). But, only distal pacing reverses the direction of propagation of the duodenal PP and its associated action potentials from aborad to orad (dotted lines).

3 March 1977 GASTRIC REFLUX BY DUODENAL PACING 431 dogs were given a 400-ml intragastric instill ate of 154 mm sodium chloride via an oral gastric tube. The tube was withdrawn as soon as the fluid had been given. The instill ate was at 22 C when administered and contained 1 % polyethylene glycol (Carbowax, Union Carbide Corporation, New York, N. Y., mean molecular weight ) as a marker. The frequencies of both the gastric and duodenal PP's were slowed 1 to 2 cycles per min by the gastric instillation of saline in every dog (P < 0.05 in every dog, Student's t-test for paired data). Accordingly, frequencies of 17 or 18 stimuli per min were used to pace the duodenal PP in these tests. This is slightly slower than the frequency of 19 to 21 cycles per min used in the pacing experiments when the stomach was empty. After 15 min, the duodenal infusion was stopped, the gastric tube was reinserted, and the stomach was aspirated and washed with 100 ml of water. These experiments were repeated 6 to 11 times in each of 4 dogs (dogs 3 through 6). A nalysis of data. At the completion of each test, the volume of the gastric aspirate and gastric wash recovered was measured. The specific activities of 5-ml samples of the original duodenal infusate, the gastric instillate, aspirate, and wash were determined in a liquid scintillation spectrometer (Packard Tri-Carb 3375, Packard Instrument Company, Inc., Downers Grove, Ill.), and the percentage of duodenal infusate refluxed into the stomach was calculated. The concentrations of polyethylene glycol in the original gastric instillate, in the aspirate, and in the wash were determined by a modification of the method of Hyden. Ii The volume of instillate emptied from the stomach in the 15-min period was then calculated by use of the following formulas: A B C and V L = 400 ml - V R v - V CA Cw R- A' C +Vw ' I C / in which V L represents the volume of the original instillate emptied from the stomach in 10 min, 400 ml is the volume originally instilled, V R is the volume ofthe original instillate remaining in the stomach at 15 min, VA is the volume of the gastric aspirate, V w is the volume of the gastric wash, and CI, C A, and C w are the concentrations of the marker in the instillate, gastric aspirate, and gastric wash, respectively. The results during no pacing, with proximal pacing, and with distal pacing were compared using Student's t-test for paired data. Results Cinefluorographic tests. Barium sulfate injected into the unstimulated duodenum was swept quickly from the duodenum into the jejunum. Little or none passed into the proximal duodenum, and on only one occasion in the eight tests did a small amount of barium enter the stomach. The refluxed barium was quickly returned to the duodenum and from there was swept distally to the jejunum. The results during proximal duodenal pacing were the same as during no pacing, except that barium never entered the stomach during any test. In contrast, most of the barium sulfate injected into the lumen of the duodenum during distal pacing passed promptly into the proximal duodenum and duodenal bulb. Little moved distally, as most of it had during no pacing or during proximal pacing. FIG. 3. Sequential cinefluorographic frames A, B, and C of gastroduodenal junction during distal duodenal pacing. An orally moving contraction of the duodenum (D) propels duodenal barium sulfate (arrow) across the pylorus (P) into the stomach (S). Orally moving duodenal contractions swept the barium against the pylorus and intermittently propelled small amounts into the stomach (fig. 3). Distal duodenal pacing caused obvious duodenal-gastric reflux in every test in every dog. The refluxed barium provided visualization of the contractions in the stomach. Their pattern appeared normal. As the barium accumulated in the stomach, the contractions became more regular and powerful, but this was probably the normal sequence of phase II contractions of the interdigestive complex. 5 The gastric contractions were nearly always peristaltic. They had their usual action in the antrum, propelling some of the barium into the duodenum and retropelling some into the gastric corpus. As distal duodenal pacing continued, a regular sequence was often established in which, with each antral cycle, some barium was forced into the duodenum with each contraction, and some was returned from the duodenum to the stomach between contractions. When distal duodenal pacing was stopped, the stomach emptied itself of the barium in the normal fashion. Then small spurts of barium passed through the pyloric canal and into the duodenum during the early phase of each peristaltic contraction, and some was

4 432 KELLY AND CODE Vol. 72, No.3 Dog no. TABLE 1. Effect of duodenal pacing on duodenal-gastric reflux While stomach was empty Mean (±SE) of duodenal counts in stomach at 15 min" (%) Proximal" Duodenal pacing Distal' During emptying of 400 ml of 154 mm NaCI Mean (±SE) of duodenal counts in stomach at 15 min a (%) Proximal" Duodenal pacing ± 9 5 9'1 9'1 34 ± ± 7 II 18" 12'i 38 ± ± d 31 d 52 d ± '1 10" 18 a SE are less than 3, except where shown. b Values differ from those with distal pacing (P < 0.05) but not from those with no pacing.,. Values differ from those with no pacing (P < 0.05). d Values differ from corresponding values when stomach was empty (P < 0.05). Distal' retropelled into the corpus of the stomach in the later phase of each antral contraction. In the absence of reflux of duodenal contents into the stomach between antral contractions, the net result was progressive gastric emptying. With cessation of distal pacing, the direction oftransit in the duodenum was almost immediately reversed and the barium present in the bulb and orad duodenum and that passing into them from the stomach was quickly cleared to the jejunum (a movie depicting the cinefluorographic findings is available from the Mayo Foundation on request). Duodenal perfusion tests. Only 1 to 3% of the radioactivity infused into the duodenum was recovered from the stomach after periods of no pacing or proximal duodenal pacing (table 1). However, with distal pacing, 15 to 40% (mean values) of the duodenal radioactivity was recovered from the stomach, showing that greatly increased duodenal-gastric reflux had occurred. Tests during gastric emptying. Much greater quantities of the duodenal infusate (mean values of9 to 34%) refluxed from the unpaced duodenum into the stomach when the stomach was emptying saline instillates compared with when it was empty (table 1). Proximal duodenal pacing did not alter the quantity of duodenalgastric reflux occurring during gastric emptying over that present with no pacing (table 1), nor did it change the rate of emptying of the gastric instill ate (table 2) or the extent of dilution of the gastric marker (table 3). In contrast, distal duodenal pacing approximately doubled duodenal-gastric reflux during gastric emptying over that occurring with no pacing or with proximal pacing. For example, mean quantities, representing 18 to 52% of the 14C activity of the polyethylene glycol injected into the duodenum, were recovered from the stomach after 15 min of distal pacing (table 1). Distal duodenal pacing also slowed the rate of emptying of the gastric instillates (table 2), although it increased the dilution of gastric marker in only 1 dog (table 3). Discussion These experiments demonstrate that the electrical and motor activity ofthe duodenum can be controlled by electrical pacing. Pacing from a site in the distal duo- TABLE 2. Effect of duodenal pacing on gastric emptying of 400 ml of 154 mmnacl Dog no. Mean (±SE) gastric instill ate emptied in 15 min (m!) During duodenal pacing Proximal Distal ± 9" 190 ± ± 17 4 II 149 ± 9" 137 ± ± 9 5 II 195 ± 13" 176 ± ± llb ± 16" 198 ± ± 1 ~ a Differs from value with distal pacing (P < 0.05) but not from value with proximal pacing. b Differs from value with proximal pacing (P < 0.05). TABLE 3. Effect of duodenal pacing on dilution of gastric marker MeanC,/C/ Dog no. During duodenum pacing Proximal Distal " 4 II II a C A and C 1 = concentrations of gastric marker in gastric aspirate and in gastric instill ate, respectively. SE = 0.02 or less. b Differs from other values in row (P < 0.05). denum reversed the direction of propagation of the duodenal PP from aborad to orad and reversed the direction of propagation of duodenal contractions and of duodenal content. Duodenal-gastric reflux resulted, and gastric emptying of gastric instillates of 154 mm NaCl was slowed. The magnitude of duodenal-gastric reflux that occurred during this experiment was no doubt greater than that documented by our tests. Our technique only measured the net amount of duodenal infusate present in the stomach at the end of a 15-min interval, not the total amount refluxed. Some of the duodenal infusate refluxed into the stomach would have been returned to the duodenum by gastric contractions during the interval. Our experiments clearly documented that distal duodenal pacing slowed the emptying of gastric instillates, but they did not quantitate the effect of pacing on the total volume of gastric content emptied. The total vol-

5 March 1977 ~ GASTRIC REFLUX BY DUODENAL PACING 433 ume of gastric content consisted not only of gastric instillate but also of secretions and duodenal content entering the gastric lumen. However, the volume of fluid other than gastric instillate entering the gastric lumen was apparently not greatly changed by pacing, because changes in the concentration of the gastric marker in the gastric content were modest (table 3). Although distal pacing approximately doubled the quantity of isotopically labeled duodenal infusate refluxed into the stomach, the total volume of infusate refluxed was too small to appreciably increase the dilution ofthe gastric marker. The amount of gastric instillate that reentered the gastric lumen from the duodenum during pacing is not known. The terminal antral segment, including the pylorus, acts in some way as a constraint to duodenal-gastric reflux. When reflux from the duodenum into the stomach was greatest, as during distal duodenal pacing, it was intermittent. It occurred between terminal antral contractions. The gastroduodenal antireflux constraint was not as effective during periods when the stomach was evacuating liquids as when the stomach was empty. This indicates that the terminal antral segment, including the pylorus, is more "open" and has more "unguarded moments" while the stomach is emptying liquids than when it is empty. The mechanism of this effect is not clear from our study. Retrograde propagation of the duodenal PP has been noted in cats during fasting and after vomiting induced by injection of morphine sulfate 7 and in dogs distal to a midduodenal transection. 3 It also may occur in human beings, but appropriate testing has not been done. REFERENCES 1. Kelly KA, Schlegel JF, Steinbach JH, et al: Duodenal-gastric reflux caused by retrograde duodenal pacing (abstr). Mayo Clin Proc 50: , Kelly KA, Steinbach JH, Schlegel JF, et al: Duodenal-gastric reflux and slowed gastric emptying by retrograde pacing of the duodenal pacesetter potential (PP). In Proceedings of the Fifth International Symposium on Gastrointestinal Motility, Herentels, Belgium, Typoff-Press, Akwari OE, Kelly KA, Steinbach JH, et al: Electric pacing of intact and transected canine small intestine and its computer model. Am J Physiol 229: , Sarna SK, Daniel EE: Electrical stimulation of small intestinal electrical control activity. Gastroenterology 69: , Code CF, Marlett JA: The interdigestive myo-electric complex of the stomach and small bowel of dogs. J Physiol (Lond) 246: , Hyden S: A turbidimetric method for the determination of higher polyethylene glycols in biological materials. K Lantbruks hogs k Ann 22: , Weisbrodt NW, Christensen J: Electrical activity of the cat duodenum in fasting and vomiting. Gastroenterology 63: , 1972