Br. J. Surg. Vol. 62 (1975) 215-220
Motility changes in the antrum after proximal gastric vagotomy MALCOLM KELLY A N D TERENCE KENNEDY* SUMMARY
The normal pattern of vesting and post-prandial motor activity in the gastric antrum has been established by observations in 6 dogs. There was a gradual increase in the amplitude of’ contraction during the first 2 hours after eating; this was maintained for 3 hours and then declined. For the first 45 minutes terminal antral contraction occurred, partially retaining and triturating the gastric contents. After 45 minutes the waves became sequential, symmetrical, increased in vigour and nctively pumped food into the duodenum. Vagotomy modified the mechanism of the antrum in various ways. Truncal and selective vagotomy reduced the work capability to 20 per cent of its normal value when recorded I month after operation. In both groups the waves were disorganized. Proximal gastric vagotomy abolished the braking mechanism and removed the initial inhibitory stimuli to antral motility. Within I month of operation the antrum had regained 58 per cent of its normal work capability and the contractions were well organized.
MOTOR denervation of the stomach is followed by a period of gastric paralysis the duration of which is unknown. This leads to gastric retention, distension and vomiting (Oberhelman and Dragstedt, 1955). Because of this difficulty the operation of vagotomy has been accompanied by a drainage procedure. In the case of pyloroplasty the delicate pyloroduodenal mechanism controlling gastric emptying is destroyed and can never be satisfactorily reconstituted; with gastro-enterostomy this region is simply bypassed. Either of these operations in combination with vagotomy converts a delicate storage and pumping mechanism controlled by fine neuro-facilitatory and inhibitory impulses into a denervated bag which empties precipitantly through a large hole in its wall. It is this transformation that produces the undesirable sequelae of dumping and diarrhoea following the operation of vagotomy and drainage. In 1957 Griffiths and Harkins suggested that gastric stasis could be avoided after vagotomy if the nerves to the antrum were preserved. The production of acid by the stomach would be reduced by dividing the branches supplying the parietal cell mass only. However, the important antral pump would remain intact. More recently, a number of centres (Amdrup and Jensen, 1970; Johnston and Wilkinson, 1970) have undertaken clinical trials of proximal gastric vagotomy, and follow-up data are accumulating, but there is no precise information about the changes in motility
of the antrum after the operation. The present study investigated the motility in the normal animal before and after truncal, selective and proximal gastric vagotomy.
Method of studying gastric motility Most of the present knowledge of gastro-intestinal motor activity is based on information from intraluminal pressure recording from balloons or opentipped manometers (Wheelon and Thomas, 1920; Hightower and Code, 1950; Creamer et al., 1956; Harris and Pope, 1964; Edwards, 1965). Both methods are subject to artefacts due to transmitted pressure waves, interference with the flow of alimentary material or obstructed tubing. They are both difficult to locate precisely. Because of these limitations a miniaturized mercury column strain gauge was developed for this investigation (Kelly et al., 1972). It consists of a fine silicone rubber tube filled with mercury under pressure (Fig. 1). As the tubing is stretched the column of mercury becomes longer and narrower, changing its electrical resistance. This in turn produces a signal which is amplified and recorded. Prior to implantation the gauges were calibrated in a waterbath at 37°C to ensure a linear response over the working range of 4-6 cm.
insulation
/ SilLcone tube
Copper electrode
Silver wire
PTFE
Self-cuing silicone rubber
I Mersilene suture
Mercury column
Mercury Silicone and copper rubber amalgam tube
Fig. 1. The mercury strain gauge; transverse section through one end of a gauge showing the junction between the mercury, the wire and the electrode.
An in vivo trial was carried out on a small number of gauges to confirm that the response was both linear and meaningful. A latex balloon was placed in the stomach and inflated with CO,. As the circumference of the antrum increased, the gauge increased in length, producing a corresponding pen deflection; this was compared with the actual measurement of the circumference and found to be directly linear (Fig. 2).
* Royal
Victoria Hospital, Belfast.
215
Malcolm Kelly and Terence Kennedy Implantation was carried out in sterile operating conditions under general anaesthesia. The gauges were accurately positioned on the serosal surface of the stomach at A1 (1 cm), A2 (2 em) and A3 (4 cm) from the pylorus (Fig. 3) and sutured in place. An adequate length of wire was left in the peritoneal cavity; the ends of the wire were taken out through a stab wound in the right flank, and then subcutaneously to the dorsal surface of the neck. Here each wire was threaded on to a needle and brought out through individual puncture holes in the skin of the neck. Vagotomy was performed at the same time as implantation, where this was appropriate.
601 50 50-
Pen
deflection
30-
0
0
10
20
30
40
50
Increase in circumference of antrurn (rnrn)
4I 60
Fig. 2. Its oiuo calibration; graph of the increase in circumference of the antrum plotted against pen deflection on the Devices M19 recorder.
Fig. 3. Pyloric antrum and first part of duodenum showing the position of the gauges.
Motility index and work capability Motility recordings were made at weekly intervals after implantation of the gauges. The results at 1 month were used for comparison as they were found to be consistent and repeatable. Recordings made in the early postoperative period usually showed a disorganized pattern of varying amplitude. The dog was given a mixed meat meal and the recordings were continued for a 7-hour post-prandial period. A motility index was obtained from the trace every 15 minutes by measuring the total excursion of gauge A2 in mm over a period of I minute. A graph representing the total post-prandial motility was then plotted from these results. In a number of the experiments the meal was mixed with barium powder and simultaneous radiological screening was performed. This enabled the gastric emptying to be directly related to the motility recording. By measuring the area under the post-prandial motility graph it was possible to obtain a numerical representation of the amount of work performed by the antrum in the 7-hour period. Although the figure
Fig. 4. Composite motility trace showing recordings from the stomach when resting. 10 minutes after eating and 45 minutes after eating. Shortly after eating, contractions at A l and A 2 occur simultaneously. A 2 preccdes A l when emptying is well established.
216
Motility changes in the antrum after vagotomy was not an absolute value of work done in foot pounds it was nevertheless useful for comparing the work capability of the stomach after the various types of vagotomy.
Results The normal stomach Terminal antral contraction : Shortly after eating it was found that the gauges at A1 and A2 contracted simultaneously (Fig. 4), indicating that the terminal antral segment had contracted as a unit. Simultaneous radiological screening studies showed that food was predominantly retained at this time (Fig. 5 ) . After 45 minutes the waves at A1 and A:! separated over the course of 5 minutes to become sequential and continued to increase in vigour. During this type of activity simultaneous radiological screening studies showed food being actively pumped into the duodenum. Posf-prandial motility contour: In the normal stomach (Fig. 6) there was a gradual build-up of motility for the first 2 hours after eating; this was maintained for 3 hours, then gradually declined. The hatched area under the graph represents the time during which terminal contraction was taking place and food was predominantly retained in the stomach. The dotted area represents sequential peristaltic activity when the food was actively pumped into the duodenum. The total area under the graph was taken as the total work capability of the normal antrum and represents 100 per cent activity. After uagotomy Nine dogs were subjected to vagotomy-3 truncal, 3 selective and 3 proximal gastric vagotomies. Strain gauges were implanted at the time o f operation. One month later an insulin test was carried out to ensure that the vagotomy was complete, and motility was recorded. After truncal vagotomy there was a dramatic decrease in overall motility and work capability of the antrum (Fig. 7). Not only was the amplitude of contraction diminished but the motility was disorganized. There were long periods of quiescence interspersed with bursts of activity (Fig. 8). This was quite unlike the regular high amplitude pattern seen in the normal stomach at this time. The total work capability was 22 per cent of normal. The changes following selective vagotomy were virtually identical (Fig. 9), the total work capability being 13 per cent of normal. When these animals were screened after a meal mixed with barium, the liquid part o f the meal was seen to leave the stomach immediately, whilst the solid part remained for many hours. After proximal gastric vagotomy the pattern was different. First, the overall motility was well maintained (Fig. 10). A most interesting feature was the speed with which high amplitude contractions occurrcd (Fig. 1 I). Within 10 minutes of eating they had reached their maximum values. In the normal stomach this was only attained 2 hours after eating. There was no initial period of terminal antral
Fig. 5. X-ray film of the dog's stomach with strain gauges A1 and A2 in situ taken 10 minutes after eating. The arrows indicate the terminal antral contraction.
Sequential peristaltic waves
Terminal antral contraction
2401 B
Motility200
Time (hr) Fig. 6. Average post-prandial motility contour of 6 normal dogs. Area under graph, 240 cm'.
320 Motility index 200 I20
Time (hr) Fig. 7. Average post-prandial motility contour of 3 dogs 1 month after truncal vagotomy. Area under graph, 54 cm?.
217
Malcolm Kelly and Terence Kennedy
Fig. 11. Motility trace from gauges A I , A2 and A3 recorded 3 weeks after proximal gastric vagotomy. It shows that contractions have reached almost maximum amplitude within 7 minutes of eating.
Fig. 8. Motility recordings from A l , A2 and A3 after truncal vagotomy showing augmented burst activity, recorded 2 hours after eating.
320280-
240 Mot iIity 2oo.
index 160. (mm)
12080-
40 0j-
4
i
Time (hr)
k
i
Fig. 9. Average post-prandial motility contour of 3 dogs 1 month after selective vagotomy. Area under graph 31 cmz.
400 3601 320
Sequential peristaltic waves
Motility
Time (hr)
Fig. 10. Average post-prandial motility contour of 3 dogs 1 month after proximal gastric vagotomy. Area under graph 138 cm2.
contraction. From the moment food was taken simultaneous radiological screening studies showed sweeping waves pumping it into the duodenum. The total work capability was 58 per cent of normal. Throughout the period of this study all the dogs in the proximal gastric vagotomy group remained in
218
good health, maintaining or even gaining weight. Following selective or truncal vagotomy the dogs did less well, losing an average of 4.0 kg. Each study was concluded at 6 months and a post-mortem examination was carried out. There was no gross change in the antrum or pyloric ring of the proximal gastric vagotomy group, but in both the selective and the truncal groups the antrum was dilated and saccular, and the circumference of the pyloric ring was consistently greater than in normal dogs.
Discussion The operation of proximal gastric vagotomy is performed without a drainage operation. It aims t o reduce gastric acidity and to allow the duodenal ulcer to heal whilst maintaining antral motility and effective gastric emptying. Does it achieve this? To confirm the effectiveness of vagotomy in this animal study, insulin stimulation tests were performed I month postoperatively. These were all negative according to Hollander’s criteria. In a pilot study of 20 patients who underwent proximal gastric vagotomy in 1970 in the Royal Victoria Hospital, Belfast, the Hollander test was found to be negative in all the patients when measured on the tenth postoperative day. This was similar to the findings of Johnston and Wilkinson (1 970). The present investigation has shown that the avoidance of a drainage operation in combination with proximal gastric vagotomy maintains an active and effective pyloric pump which recovers rapidly after operation. This was in sharp contrast to the almost complete absence of motor activity seen after truncal and selective vagotomy. Stasis is unlikely to occur after proximal gastric vagotomy. Without a drainage procedure the pyloric ring remains intact and will completely close at the end of each peristaltic cycle. This was first shown by Hofmeister and Schutz (1886) and by Moritz and Tschr (1901). It was confirmed by Cole (1913) and studied in detail by Quigley and Meschan (1941). The
Motility changes in the antrum after vagotomy contracted pylorus not only provides a ‘breach block’ for duodenal contraction, but prevents the regurgitation of bile, a common problem after either pyloroplasty or gastrojejunostomy. It was Cannon in 1898 who first pointed out that the pyloric region of the stomach had a more important function than that of merely expelling gastric contents into the intestine. He found that adding a hard pellet of starch to a gruel and bismuth meal given to a cat delayed gastric emptying from 15 to 45 minutes. Carlson et al. (l966), using cineradiographic techniques, described the mechanism through which this delay in gastric emptying was brought about. When peristaltic waves reach the antrum they produce a diaphragm which may or may not completely separate the antrum from the body of the stomach. At this moment a contraction involving the whole of the terminal antral segment, including the pyloric sphincter, occurs. Alimentary contents will be forcefully ejected into the duodenum until such time as the pyloric sphincter closes and prevents this flow. Since the pyloric canal is much narrower than the rest of the antrum it remains closed throughout the continuation of the terminal antral contraction. At other times the pyloric canal will narrow early in the cycle but not completely close so that movement of the contents through it into the duodenum occurs while the antrum continues to contract. This agrees with Hunt’s (Hunt and Knox, 1968) theory of the regulation of gastric emptying, depending on the timing of the closure of the pylorus, during antral systole. An unexpected finding in the present study was the complete absence of terminal antral contractionand the ‘antral braking mechanism’-after all types of vagotomy. Following a normal meal there is an initial period during which very little emptying occurs from the dog’s stomach. The food is held back by a combination of contractions involving the terminal 3 cm of antrum, high antral tone and low amplitude peristaltic waves. The result of this is retropulsion of food into the body of the stomach. This capability is lost after vagotomy. After proximal gastric vagotomy high amplitude peristaltic waves occur within minutes of eating, pumping the food into the duodenum with the result that a hypertonic meal could be delivered into the duodenum before equilibration, producing dumping. This finding has been confirmed in man by Clarke and Alexander-Williams (1973). They found that the initial emptying of a f u i d meal was faster after proximal gastric vagotomy than in the normal individual. Dumping has been reported in various clinical trials (Humphrey et al., 1972; Clarke and Alexander-Williams, 1973). In our own pilot study 40 per cent of the patients admitted to mild transitory dumping at some time during the early postoperative period, but at 1 year after operation none had significant dumping symptoms. The study has also thrown light on the phenomenon of dumping after the more conventional truncal and selective vagotomy. Again the braking mechanism was found to be completely absent after operation.
Both radiological screening studies and post-mortem findings showed the antrum to be dilated and saccular, and the circumference of the pyloric ring to be consistently greater than in normal animals. It might be said that the denervated stomach behaves like the neurogenic bladder; if allowed to overdistend, as in the case of truncal and selective vagotomy without drainage, it becomes a dilated bag which empties by overflow. Although the finer mechanisms of the antrum are disorganized after proximal gastric vagotomy, vigorous peristaltic activity remains, emptying is satisfactory and it is therefore unnecessary t o perform an accompanying drainage operation.
Acknowledgements This work was supported by a grant from the Royal Victoria Hospital, Belfast, and was part of a thesis for the Mastership in Surgery of London University in 1973. References and JENSEN H. E. (1970) Selective vagotomy of the parietal cell mass preserving innervation of the undrained antrum: a preliminary report of results in patients with duodenal ulcer. Castroenterology 59, 522-521. CANNON W. B. (1898) Movements Of the stomach studied by roentgen rays. Am. J . Physiol. 1, 359. CARLSON H. C., CODE C. and NELSON R. A. (1966) Motor action of the canine gastroduodenal junction. A cineradiographic pressure and electrical study. Am. J. Dig. Dis. 11, 155-172. CLARKE R. J . and ALEXANDER-WILLIAMS J . (1973) The effect of preserving antral innervation and of a pyloroplasty on gastric emptying after vagotomy in man. Gut 14, 300-307. COLE L. G . (1913) Physiology of the pylorus, pilleus ventriculi and duodenum as observed roentgenographically. J A M A 61, 762. CREAMER B., ANDERSEN H. A. and CODE c . F. (1956) Oesophageal motility in patients with scleroderma and related disease. Castroenteroligia 86, 763. EDWARDS D. A. w . (1965) Techniques and interpretationsof measurement of gastrointestinal pressures. Ann. R . Coll. Surg. Engl. 31, 215-284. GRIFFITHS c . A. and HARKINS H. N. (1957) Partial gastric vagotomy-an experimental study. Castroenterology 32, 96. HARRIS L. D. and POPE c . E. (1964) “Squeeze” versus resistance: an evaluation of the mechanism of sphincter competence. J . Clin. Invest. 43, 2272-2278. HIGHTOWERN . c . and CODE C. F. (1950) The quantitative analysis of antral gastric motility records in normal human beings, with a study of the effects of neostigmine. Proc. Staff Meet. Muyo Clin. 25, 691-704. HOFMEISTER F. and SCHUTZ E. ( I 886) Ueber Resorption und Assimilation der Nahrstoffe. Arch. Exp. Path. AMDRUP E.
Phnrmnk. 20, 7. 219
Malcolm Kelly and Terence Kennedy
c. s., WILKINSON A. R. and JOHNSTON D. (1972) The correlation between the rate of gastric emptying and dumping and diarrhoea after truncal, selective, and highly selective vagotomy. Br. J. Surg. 59, 309. HUNT J. N. and KNOX M. T. (1968) Control of gastric emptying. Am. J. Dig. Dis. 13, 372-375. JOHNSTON D. and WILKINSON A. R. (1970) Highly selective vagotorny without a drainage procedure in the treatment of duodenal ulcer. Br. J. Surg. 57, 289-296. KELLY J. M., CONNELL A. M., KENNEDY T. L. and CLARKE E. w. (1972) Application o f miniaturized mercury column strain gauges to the study of gastric motility. Am. J. Dig. Dis. 17, 455-462. MORITZ s. and TSCHR z. (1901) Ueber die Beziehungen zwischen Arzneien und Magen. Biol. 42 (NS 24), 565. HUMPHREY
220
M U R R A Y J. G. (1969) Regeneration In: ALEXANDER-WILLIAMS J.
of the vagus nerve. and cox A. G. (ed.) After Vagotomy. London, Butterworths, pp. 77-88. OBERHELMAN H. A. and DRAGSTEDT L. R. (1955) New physiologic concepts related to the surgical treatment of duodenal ulcer by vagotomy and gastroenterostomy. Surg. Cynecol. Obstet. 101, 194-200. QUICLEY J. P. and MESCHAN I. (1941) Inhibition of the pyloric sphincter region by the digestive products of fat. Am. J . Physiol. 134, 803-807. WHEELON H. and THOMAS J. E. (1920) Observations on the motility o f the antrum and the relation of the rhythmic activity of the pyloric sphincter to that of the duodenum. J . Lab. Clin. Med. 6, 124.
Motility changes in the antrum after proximal gastric vagotomy MALCOLM KELLY A N D TERENCE KENNEDY* SUMMARY
The normal pattern of vesting and post-prandial motor activity in the gastric antrum has been established by observations in 6 dogs. There was a gradual increase in the amplitude of’ contraction during the first 2 hours after eating; this was maintained for 3 hours and then declined. For the first 45 minutes terminal antral contraction occurred, partially retaining and triturating the gastric contents. After 45 minutes the waves became sequential, symmetrical, increased in vigour and nctively pumped food into the duodenum. Vagotomy modified the mechanism of the antrum in various ways. Truncal and selective vagotomy reduced the work capability to 20 per cent of its normal value when recorded I month after operation. In both groups the waves were disorganized. Proximal gastric vagotomy abolished the braking mechanism and removed the initial inhibitory stimuli to antral motility. Within I month of operation the antrum had regained 58 per cent of its normal work capability and the contractions were well organized.
MOTOR denervation of the stomach is followed by a period of gastric paralysis the duration of which is unknown. This leads to gastric retention, distension and vomiting (Oberhelman and Dragstedt, 1955). Because of this difficulty the operation of vagotomy has been accompanied by a drainage procedure. In the case of pyloroplasty the delicate pyloroduodenal mechanism controlling gastric emptying is destroyed and can never be satisfactorily reconstituted; with gastro-enterostomy this region is simply bypassed. Either of these operations in combination with vagotomy converts a delicate storage and pumping mechanism controlled by fine neuro-facilitatory and inhibitory impulses into a denervated bag which empties precipitantly through a large hole in its wall. It is this transformation that produces the undesirable sequelae of dumping and diarrhoea following the operation of vagotomy and drainage. In 1957 Griffiths and Harkins suggested that gastric stasis could be avoided after vagotomy if the nerves to the antrum were preserved. The production of acid by the stomach would be reduced by dividing the branches supplying the parietal cell mass only. However, the important antral pump would remain intact. More recently, a number of centres (Amdrup and Jensen, 1970; Johnston and Wilkinson, 1970) have undertaken clinical trials of proximal gastric vagotomy, and follow-up data are accumulating, but there is no precise information about the changes in motility
of the antrum after the operation. The present study investigated the motility in the normal animal before and after truncal, selective and proximal gastric vagotomy.
Method of studying gastric motility Most of the present knowledge of gastro-intestinal motor activity is based on information from intraluminal pressure recording from balloons or opentipped manometers (Wheelon and Thomas, 1920; Hightower and Code, 1950; Creamer et al., 1956; Harris and Pope, 1964; Edwards, 1965). Both methods are subject to artefacts due to transmitted pressure waves, interference with the flow of alimentary material or obstructed tubing. They are both difficult to locate precisely. Because of these limitations a miniaturized mercury column strain gauge was developed for this investigation (Kelly et al., 1972). It consists of a fine silicone rubber tube filled with mercury under pressure (Fig. 1). As the tubing is stretched the column of mercury becomes longer and narrower, changing its electrical resistance. This in turn produces a signal which is amplified and recorded. Prior to implantation the gauges were calibrated in a waterbath at 37°C to ensure a linear response over the working range of 4-6 cm.
insulation
/ SilLcone tube
Copper electrode
Silver wire
PTFE
Self-cuing silicone rubber
I Mersilene suture
Mercury column
Mercury Silicone and copper rubber amalgam tube
Fig. 1. The mercury strain gauge; transverse section through one end of a gauge showing the junction between the mercury, the wire and the electrode.
An in vivo trial was carried out on a small number of gauges to confirm that the response was both linear and meaningful. A latex balloon was placed in the stomach and inflated with CO,. As the circumference of the antrum increased, the gauge increased in length, producing a corresponding pen deflection; this was compared with the actual measurement of the circumference and found to be directly linear (Fig. 2).
* Royal
Victoria Hospital, Belfast.
215
Malcolm Kelly and Terence Kennedy Implantation was carried out in sterile operating conditions under general anaesthesia. The gauges were accurately positioned on the serosal surface of the stomach at A1 (1 cm), A2 (2 em) and A3 (4 cm) from the pylorus (Fig. 3) and sutured in place. An adequate length of wire was left in the peritoneal cavity; the ends of the wire were taken out through a stab wound in the right flank, and then subcutaneously to the dorsal surface of the neck. Here each wire was threaded on to a needle and brought out through individual puncture holes in the skin of the neck. Vagotomy was performed at the same time as implantation, where this was appropriate.
601 50 50-
Pen
deflection
30-
0
0
10
20
30
40
50
Increase in circumference of antrurn (rnrn)
4I 60
Fig. 2. Its oiuo calibration; graph of the increase in circumference of the antrum plotted against pen deflection on the Devices M19 recorder.
Fig. 3. Pyloric antrum and first part of duodenum showing the position of the gauges.
Motility index and work capability Motility recordings were made at weekly intervals after implantation of the gauges. The results at 1 month were used for comparison as they were found to be consistent and repeatable. Recordings made in the early postoperative period usually showed a disorganized pattern of varying amplitude. The dog was given a mixed meat meal and the recordings were continued for a 7-hour post-prandial period. A motility index was obtained from the trace every 15 minutes by measuring the total excursion of gauge A2 in mm over a period of I minute. A graph representing the total post-prandial motility was then plotted from these results. In a number of the experiments the meal was mixed with barium powder and simultaneous radiological screening was performed. This enabled the gastric emptying to be directly related to the motility recording. By measuring the area under the post-prandial motility graph it was possible to obtain a numerical representation of the amount of work performed by the antrum in the 7-hour period. Although the figure
Fig. 4. Composite motility trace showing recordings from the stomach when resting. 10 minutes after eating and 45 minutes after eating. Shortly after eating, contractions at A l and A 2 occur simultaneously. A 2 preccdes A l when emptying is well established.
216
Motility changes in the antrum after vagotomy was not an absolute value of work done in foot pounds it was nevertheless useful for comparing the work capability of the stomach after the various types of vagotomy.
Results The normal stomach Terminal antral contraction : Shortly after eating it was found that the gauges at A1 and A2 contracted simultaneously (Fig. 4), indicating that the terminal antral segment had contracted as a unit. Simultaneous radiological screening studies showed that food was predominantly retained at this time (Fig. 5 ) . After 45 minutes the waves at A1 and A:! separated over the course of 5 minutes to become sequential and continued to increase in vigour. During this type of activity simultaneous radiological screening studies showed food being actively pumped into the duodenum. Posf-prandial motility contour: In the normal stomach (Fig. 6) there was a gradual build-up of motility for the first 2 hours after eating; this was maintained for 3 hours, then gradually declined. The hatched area under the graph represents the time during which terminal contraction was taking place and food was predominantly retained in the stomach. The dotted area represents sequential peristaltic activity when the food was actively pumped into the duodenum. The total area under the graph was taken as the total work capability of the normal antrum and represents 100 per cent activity. After uagotomy Nine dogs were subjected to vagotomy-3 truncal, 3 selective and 3 proximal gastric vagotomies. Strain gauges were implanted at the time o f operation. One month later an insulin test was carried out to ensure that the vagotomy was complete, and motility was recorded. After truncal vagotomy there was a dramatic decrease in overall motility and work capability of the antrum (Fig. 7). Not only was the amplitude of contraction diminished but the motility was disorganized. There were long periods of quiescence interspersed with bursts of activity (Fig. 8). This was quite unlike the regular high amplitude pattern seen in the normal stomach at this time. The total work capability was 22 per cent of normal. The changes following selective vagotomy were virtually identical (Fig. 9), the total work capability being 13 per cent of normal. When these animals were screened after a meal mixed with barium, the liquid part o f the meal was seen to leave the stomach immediately, whilst the solid part remained for many hours. After proximal gastric vagotomy the pattern was different. First, the overall motility was well maintained (Fig. 10). A most interesting feature was the speed with which high amplitude contractions occurrcd (Fig. 1 I). Within 10 minutes of eating they had reached their maximum values. In the normal stomach this was only attained 2 hours after eating. There was no initial period of terminal antral
Fig. 5. X-ray film of the dog's stomach with strain gauges A1 and A2 in situ taken 10 minutes after eating. The arrows indicate the terminal antral contraction.
Sequential peristaltic waves
Terminal antral contraction
2401 B
Motility200
Time (hr) Fig. 6. Average post-prandial motility contour of 6 normal dogs. Area under graph, 240 cm'.
320 Motility index 200 I20
Time (hr) Fig. 7. Average post-prandial motility contour of 3 dogs 1 month after truncal vagotomy. Area under graph, 54 cm?.
217
Malcolm Kelly and Terence Kennedy
Fig. 11. Motility trace from gauges A I , A2 and A3 recorded 3 weeks after proximal gastric vagotomy. It shows that contractions have reached almost maximum amplitude within 7 minutes of eating.
Fig. 8. Motility recordings from A l , A2 and A3 after truncal vagotomy showing augmented burst activity, recorded 2 hours after eating.
320280-
240 Mot iIity 2oo.
index 160. (mm)
12080-
40 0j-
4
i
Time (hr)
k
i
Fig. 9. Average post-prandial motility contour of 3 dogs 1 month after selective vagotomy. Area under graph 31 cmz.
400 3601 320
Sequential peristaltic waves
Motility
Time (hr)
Fig. 10. Average post-prandial motility contour of 3 dogs 1 month after proximal gastric vagotomy. Area under graph 138 cm2.
contraction. From the moment food was taken simultaneous radiological screening studies showed sweeping waves pumping it into the duodenum. The total work capability was 58 per cent of normal. Throughout the period of this study all the dogs in the proximal gastric vagotomy group remained in
218
good health, maintaining or even gaining weight. Following selective or truncal vagotomy the dogs did less well, losing an average of 4.0 kg. Each study was concluded at 6 months and a post-mortem examination was carried out. There was no gross change in the antrum or pyloric ring of the proximal gastric vagotomy group, but in both the selective and the truncal groups the antrum was dilated and saccular, and the circumference of the pyloric ring was consistently greater than in normal dogs.
Discussion The operation of proximal gastric vagotomy is performed without a drainage operation. It aims t o reduce gastric acidity and to allow the duodenal ulcer to heal whilst maintaining antral motility and effective gastric emptying. Does it achieve this? To confirm the effectiveness of vagotomy in this animal study, insulin stimulation tests were performed I month postoperatively. These were all negative according to Hollander’s criteria. In a pilot study of 20 patients who underwent proximal gastric vagotomy in 1970 in the Royal Victoria Hospital, Belfast, the Hollander test was found to be negative in all the patients when measured on the tenth postoperative day. This was similar to the findings of Johnston and Wilkinson (1 970). The present investigation has shown that the avoidance of a drainage operation in combination with proximal gastric vagotomy maintains an active and effective pyloric pump which recovers rapidly after operation. This was in sharp contrast to the almost complete absence of motor activity seen after truncal and selective vagotomy. Stasis is unlikely to occur after proximal gastric vagotomy. Without a drainage procedure the pyloric ring remains intact and will completely close at the end of each peristaltic cycle. This was first shown by Hofmeister and Schutz (1886) and by Moritz and Tschr (1901). It was confirmed by Cole (1913) and studied in detail by Quigley and Meschan (1941). The
Motility changes in the antrum after vagotomy contracted pylorus not only provides a ‘breach block’ for duodenal contraction, but prevents the regurgitation of bile, a common problem after either pyloroplasty or gastrojejunostomy. It was Cannon in 1898 who first pointed out that the pyloric region of the stomach had a more important function than that of merely expelling gastric contents into the intestine. He found that adding a hard pellet of starch to a gruel and bismuth meal given to a cat delayed gastric emptying from 15 to 45 minutes. Carlson et al. (l966), using cineradiographic techniques, described the mechanism through which this delay in gastric emptying was brought about. When peristaltic waves reach the antrum they produce a diaphragm which may or may not completely separate the antrum from the body of the stomach. At this moment a contraction involving the whole of the terminal antral segment, including the pyloric sphincter, occurs. Alimentary contents will be forcefully ejected into the duodenum until such time as the pyloric sphincter closes and prevents this flow. Since the pyloric canal is much narrower than the rest of the antrum it remains closed throughout the continuation of the terminal antral contraction. At other times the pyloric canal will narrow early in the cycle but not completely close so that movement of the contents through it into the duodenum occurs while the antrum continues to contract. This agrees with Hunt’s (Hunt and Knox, 1968) theory of the regulation of gastric emptying, depending on the timing of the closure of the pylorus, during antral systole. An unexpected finding in the present study was the complete absence of terminal antral contractionand the ‘antral braking mechanism’-after all types of vagotomy. Following a normal meal there is an initial period during which very little emptying occurs from the dog’s stomach. The food is held back by a combination of contractions involving the terminal 3 cm of antrum, high antral tone and low amplitude peristaltic waves. The result of this is retropulsion of food into the body of the stomach. This capability is lost after vagotomy. After proximal gastric vagotomy high amplitude peristaltic waves occur within minutes of eating, pumping the food into the duodenum with the result that a hypertonic meal could be delivered into the duodenum before equilibration, producing dumping. This finding has been confirmed in man by Clarke and Alexander-Williams (1973). They found that the initial emptying of a f u i d meal was faster after proximal gastric vagotomy than in the normal individual. Dumping has been reported in various clinical trials (Humphrey et al., 1972; Clarke and Alexander-Williams, 1973). In our own pilot study 40 per cent of the patients admitted to mild transitory dumping at some time during the early postoperative period, but at 1 year after operation none had significant dumping symptoms. The study has also thrown light on the phenomenon of dumping after the more conventional truncal and selective vagotomy. Again the braking mechanism was found to be completely absent after operation.
Both radiological screening studies and post-mortem findings showed the antrum to be dilated and saccular, and the circumference of the pyloric ring to be consistently greater than in normal animals. It might be said that the denervated stomach behaves like the neurogenic bladder; if allowed to overdistend, as in the case of truncal and selective vagotomy without drainage, it becomes a dilated bag which empties by overflow. Although the finer mechanisms of the antrum are disorganized after proximal gastric vagotomy, vigorous peristaltic activity remains, emptying is satisfactory and it is therefore unnecessary t o perform an accompanying drainage operation.
Acknowledgements This work was supported by a grant from the Royal Victoria Hospital, Belfast, and was part of a thesis for the Mastership in Surgery of London University in 1973. References and JENSEN H. E. (1970) Selective vagotomy of the parietal cell mass preserving innervation of the undrained antrum: a preliminary report of results in patients with duodenal ulcer. Castroenterology 59, 522-521. CANNON W. B. (1898) Movements Of the stomach studied by roentgen rays. Am. J . Physiol. 1, 359. CARLSON H. C., CODE C. and NELSON R. A. (1966) Motor action of the canine gastroduodenal junction. A cineradiographic pressure and electrical study. Am. J. Dig. Dis. 11, 155-172. CLARKE R. J . and ALEXANDER-WILLIAMS J . (1973) The effect of preserving antral innervation and of a pyloroplasty on gastric emptying after vagotomy in man. Gut 14, 300-307. COLE L. G . (1913) Physiology of the pylorus, pilleus ventriculi and duodenum as observed roentgenographically. J A M A 61, 762. CREAMER B., ANDERSEN H. A. and CODE c . F. (1956) Oesophageal motility in patients with scleroderma and related disease. Castroenteroligia 86, 763. EDWARDS D. A. w . (1965) Techniques and interpretationsof measurement of gastrointestinal pressures. Ann. R . Coll. Surg. Engl. 31, 215-284. GRIFFITHS c . A. and HARKINS H. N. (1957) Partial gastric vagotomy-an experimental study. Castroenterology 32, 96. HARRIS L. D. and POPE c . E. (1964) “Squeeze” versus resistance: an evaluation of the mechanism of sphincter competence. J . Clin. Invest. 43, 2272-2278. HIGHTOWERN . c . and CODE C. F. (1950) The quantitative analysis of antral gastric motility records in normal human beings, with a study of the effects of neostigmine. Proc. Staff Meet. Muyo Clin. 25, 691-704. HOFMEISTER F. and SCHUTZ E. ( I 886) Ueber Resorption und Assimilation der Nahrstoffe. Arch. Exp. Path. AMDRUP E.
Phnrmnk. 20, 7. 219
Malcolm Kelly and Terence Kennedy
c. s., WILKINSON A. R. and JOHNSTON D. (1972) The correlation between the rate of gastric emptying and dumping and diarrhoea after truncal, selective, and highly selective vagotomy. Br. J. Surg. 59, 309. HUNT J. N. and KNOX M. T. (1968) Control of gastric emptying. Am. J. Dig. Dis. 13, 372-375. JOHNSTON D. and WILKINSON A. R. (1970) Highly selective vagotorny without a drainage procedure in the treatment of duodenal ulcer. Br. J. Surg. 57, 289-296. KELLY J. M., CONNELL A. M., KENNEDY T. L. and CLARKE E. w. (1972) Application o f miniaturized mercury column strain gauges to the study of gastric motility. Am. J. Dig. Dis. 17, 455-462. MORITZ s. and TSCHR z. (1901) Ueber die Beziehungen zwischen Arzneien und Magen. Biol. 42 (NS 24), 565. HUMPHREY
220
M U R R A Y J. G. (1969) Regeneration In: ALEXANDER-WILLIAMS J.
of the vagus nerve. and cox A. G. (ed.) After Vagotomy. London, Butterworths, pp. 77-88. OBERHELMAN H. A. and DRAGSTEDT L. R. (1955) New physiologic concepts related to the surgical treatment of duodenal ulcer by vagotomy and gastroenterostomy. Surg. Cynecol. Obstet. 101, 194-200. QUICLEY J. P. and MESCHAN I. (1941) Inhibition of the pyloric sphincter region by the digestive products of fat. Am. J . Physiol. 134, 803-807. WHEELON H. and THOMAS J. E. (1920) Observations on the motility o f the antrum and the relation of the rhythmic activity of the pyloric sphincter to that of the duodenum. J . Lab. Clin. Med. 6, 124.
