Br. J. Surg. 1991, Vol. 78, January, 2427
M. N. Hartley and C. R. Mackie University Department of Surgery, Royal Liverpool Hospital, Liverpool 1 7 BXP, UK Correspondence to: Mr C . R. Mackie, Department of
Surgery, Royal Liverpool Hospital, PO Box 147, Liverpool L69 3BX, UK
Gastric adaptive relaxation and symptoms after vagotomy Gastric adaptive relaxation is reported to be impaired after vagotomy. This abnormality has been implicated in the pathogenesis of postvagotomy symptoms, but no association has previously been demonstrated between the two. Gastric adaptive relaxation was measured in 15 healthy volunteers and 33 patients more than I year after highly selective vagotomy or truncal vagotomy and drainage. Seventeen patients were asymptomatic. The remainder were symptomatic including seven patients with persistent diarrhoea. Fasted subjects were intubated with a Ryie 's tube containing a pressure microtransducer within a flaccid (800 ml) plastic bag. Gastric corpus-fundus pressure was recorded during distension of the bag with air (15 mils) over 30 s. Pressure indices were derived f r o m the areas under the pressure curves. Median (range) pressure indices were: healthy volunteers 12.7 (7.5-1 7.1) cmH,O, highly selective vagotomy 14.0 (9.8-15.9) cmH,O (n.s.), truncal vagotomy and drainage 14.5 (8.6-26.8) cmH,O (P = 0.04). All patients with diarrhoea had abnormally high pressure indices (P < 0.001). Pressure indices in all' other patient groups were within the normal range. W e conclude that gastric adaptive relaxation remains abnormal in patients with postvagotomy diarrhoea but not in those who are asymptomatic or who have other symptoms.
Impairment of gastric adaptive relaxation has been reported after both truncal ~ a g o t o m y ' -and ~ highly selective vagotomy6' and the degree of impairment is said to be similar after either ~ p e r a t i o n ' . ~ Studies . in both man and animals have been carried out after vagotomy, at l a p a r ~ t o m y ' - ~and ~ a few months later'.2*4*6.8*'2 . Few studies, however, have measured gastric adaptive relaxation more than 1 year after surgery. The impairment of gastric adaptive relaxation, in combination with other motility disturbances after vagotomy, is implicated in the pathophysiology of postvagotomy symptoms. Symptoms after truncal vagotomy and drainage tend to be worse than after highly selective vagotomy. This has been attributed to the more disturbed gastric emptying associated with the drainage procedure and denervation of the antropyloric segment and the small bowe1'3*'4. It is recognized, for example, that diarrhoea after highly selective vagotomy is uncommon, while after truncal vagotomy and drainage diarrhoea is frequently reported and may be a serious and persistent problem for a minority of patient^'^. The contribution of impaired gastric adaptive relaxation to the pathogenesis of postvagotomy disorders is both speculative and controversial. Postprandial epigastric fullness has been attributed to the failure of gastric adaptive relaxationis4. Early rapid gastric emptying is thought to be the key abnormality causing the dumping syndrome and may also contribute to the pathogenesis of postvagotomy diarrhoea. Early rapid gastric emptying is thought partly to be due to impaired gastric adaptive relaxation*. There are, therefore, clinically recognized links between impairment of gastric adaptive relaxation and the presence of symptoms after vagotomy. To date, there has been no study of gastric adaptive relaxation specifically comparing patients with and without symptoms after vagotomy. The aim of this study was to measure gastric adaptive relaxation in patients after highly selective vagotomy and truncal vagotomy and drainage in relation to their various symptoms more than 1 year after surgery. The study was formally approved by the Royal Liverpool Hospital
24
Ethical Committee in Novlember 1985. Written informed consent was given by all subjects studied.
Patients and methods Fifteen healthy volunteers were studied (13 men and two women). Their median age and body-weight were 30 (range 2 2 4 1 ) years and 70 (range 5G79) kg. All healthy volunteers 'were asymptomaticon no medication. None had undergone any previous upper gastrointestiiial surgery. There were 33 patients in the study. All had previously undergone vagotomy for intractable duodenal ulceration, confirmed by endoscopy or barium meal and subsequently at laparotomy, or carried out for complications of duodenal ulcer. Twenty-five of the patients (20 men and five women) had had trilncal vagotomy and drainage (18 pyloroplasty and seven gastroenterostomy) and eight patients, highly selective vagotomy (six men and two women). The median age and body-weight for patients in the truncal vagotomy and drainage group were 51 (range 35-67) years and 63 (range 44-92) kg. The median age and body-weight for patients in the highly selective vagotomy group were 35 (range 2 3 4 7 ) years and '77 (range 59-91) kg. The median time since operation was 4 years 6 months (range 1-24 years) for patients after truncal vagotomy and drainage and 2 (range 1-3) years for patients after highly selective vagotomy. Five patients after highly selective vagotomy and 12 patients after truncal vagotomy and drainage were asymptomatic. Symptoms in the remaining patients, summarized in Table 1 , had been present for over 1 year. Some symptomatic patients had more than one symptom. Table 2 shows the nature and si:verity of diarrhoea in patients with this symptom. No patient had or gave a history of recurrent ulceration following vagotomy. All symptomatic patients had undergone routine investigationof their symptoms. All patients studied were systematically classified in terms of presence or absence, nature and severity of symptoms before undergoing the studies described. The healthy volunteer and patient groups were studied in the same manner. The technique for measuring gastric adaptive relaxation was a modification of that used by Jahnberg et al.3 and identical to that recently reported by our groupi6. After an overnight fast, subjects swallowed, via the nose, a 10 Fr Ryle's tube with a collapsed plastic bag (800ml) sealed over the end, the side holes of the tube being within the bag. Within the lumen of the tube, positioned near the tip, was a
0007-1323/9l/OlOO24-04
(i? 1991 Butterworth-Helnemann Ltd
Gastric adaptive relaxation and symptoms after vagotomy: M . N. Hartley and C. R. Mackie
Table 1 Symptoms in patients after vagotomy
Highly selective vagotomy Truncal vagotomy and drainage
No. of patients
Symptoms
5 3 12 5 3 2 I
None Epigastric fullness None Epigastric fullness Bilious vomiting Vasomotor dumping Diarrhoea
median pressure index of 13.8 cmH,O (Figure 3). The median pressure index for the symptomatic patients after truncal vagotomy was 17.6 cmH,O and this was significantly higher ( P < 0.025) than both healthy volunteers and asymptomatic patients. Further analysis of symptoms in relation to pressure indices
....................................
Table 2 Characteristics of postvagotomy diarrhoea in seven patients after truncal vagotomy and drainage
180
In s u f f l a t i o n
Diarrhoea Continuous/episodic
Stool frequency (maximum per day)
Time I s )
Urgency Dynamic pressure index = Area under curve during insufflation
+
Continuous Continuous Continuous Continuous Episodic Episodic Episodic
After insufflation
+ +++ +++ +++ + ++
Continuous, loose stools every day; episodic, loose stools at intervals of days or weeks; urgency: +, urge to defaecate; + + , precipitate urge to defaecate; + +, incontinent of faeces
c m ~ Z ~
30 ’
Figure 1 Derivation of the dynamic pressure index from the pressure recording
30
+
pressure microtransducer (Gaeltec, Dingwall, Rosshire, UK) which was calibrated against a water manometer to provide a full scale deflection of 50 cm of water. The transducer was interfaced with an amplifier and pen recorder (Could Electronics Ltd., Coventry, UK) to record pressure changes against time. The bag was positioned manometrically in the corpus-fundus of the stomach. A baseline recording was obtained during a 15 min rest period. Air from a compressed air cylinder (British Oxygen Company, Guildford, UK) was delivered at a constant rate via an air flow meter and a safety valve to the Ryle’s tube. This constant flow of air was used to distend the hag at a rate of 15 ml/s for 30 s during which the pressure rise within the intragastric bag was recorded. The recording was continued at this end volume for a further 150 s. The air was then aspirated with a syringe and the volume noted. This procedure was repeated at 15 min intervals to provide four recordings for each subject, two each with subjects sitting and supine. For each recording a dynamic pressure index was derived from the area under the curve during distension. These measurements were ascribed units in cmH,O (Figure I). A single value was then derived for each subject by taking the mean of the four observations. Statistical analyses of pressure indices between groups were made using the Kruskal-Wallis and Mann-Whitney I/ tests. Analysis of variance was used to compare volumes of air aspirated between groups. The 1’ test was used to compare the types of drainage procedure.
25
za
1!
0
: I
0
l 0
1(
0
Results As previously reported” there was no identifiable trend of variation in successive recordings during individual studies. There was no detectable difference between sitting and supine recordings. All pressure indices in patients after highly selective vagotomy were within the range obtained for healthy volunteers (Figure 2). The median pressure index for these patients (14.0 cmH,O) was higher than that for the healthy volunteer group (12.7 cmH,O) but this difference was not statistically significant. Pressure indices in patients after truncal vagotomy and drainage were significantly higher than those in the healthy volunteer group (P=0.04). The median pressure index for patients after truncal vagotomy and drainage was 14.5 cmH,O (Figure 2). When these patients were divided into asymptomatic and symptomatic groups, all of the asymptomatic patients had pressure indices within the range for healthy volunteers with a
Br. J. Surg.. Vol. 78, No. 1, January 1991
I Controls
I Highly selective vagotorny
I T r u n c a l vagotorny a n d drainage
Figure 2 Scattergram of pressure indices in healthy volunteers and patienls after highly selective vagotomy and truncal vagotomy and drainage. Median values shown by bars. ( 0 ) Asymptomatic; (H) symptomatic
25
Gastric adaptive relaxation and symptoms after vagotomy: M. N. Hartley and C. R. Mackie
different (51 years (median; range 39-67 years) and 65 (44-92) kg respectively). There was no statistical correlation between the type of drainage procedure and the presence or absence of postvagotomy diarrhoea; four of 18 patients without diarrhoea and three of seven patients with diarrhoea had a gastroenterostomy.
30
25
Discussion
ON
I
-5 X
g .-C
+
2c
1:
a
. I . .. . .
-c
3
"1 Lo
L
CL
11
t
8
I
Control
.. . .. . .
8
8
-
. . -
I I I Asymptomatic Symptomatic T r u n c a l t r unca I truncal vagotorny vagotorny vagotorny and drainage a n d d r a in a g e a n d d r a in a g e ( d i a r r h o e a )
Figure 3 Scattergram of pressure indices in healthy volunteers and asymptomatic and symptomatic patients afer truncal vagotomy and drainage. Patients with postoagotomy diarrhoea are shown separately. Median values shown by bars. (e)Asymptomatic; (m) symptomatic
revealed that patients with epigastric fullness tended to have higher pressure indices in both vagotomy groups; three patients with epigastric fullness in the highly selective vagotomy group were within the top four pressure indices of their group (Figure 2 ) and four of the five patients with epigastric fullness after truncal vagotomy had pressure indices above the range for healthy volunteers. The latter patients also had postvagotomy diarrhoea. Patients with bilious vomiting and vasomotor dumping had pressure indices within the range for healthy volunteers. All seven patients with postvagotomy diarrhoea had pressure induces above the range for healthy volunteers. The median pressure index for the patients in this group was 20.3 cmH,O. This value was significantly higher (P
M. N. Hartley and C. R. Mackie University Department of Surgery, Royal Liverpool Hospital, Liverpool 1 7 BXP, UK Correspondence to: Mr C . R. Mackie, Department of
Surgery, Royal Liverpool Hospital, PO Box 147, Liverpool L69 3BX, UK
Gastric adaptive relaxation and symptoms after vagotomy Gastric adaptive relaxation is reported to be impaired after vagotomy. This abnormality has been implicated in the pathogenesis of postvagotomy symptoms, but no association has previously been demonstrated between the two. Gastric adaptive relaxation was measured in 15 healthy volunteers and 33 patients more than I year after highly selective vagotomy or truncal vagotomy and drainage. Seventeen patients were asymptomatic. The remainder were symptomatic including seven patients with persistent diarrhoea. Fasted subjects were intubated with a Ryie 's tube containing a pressure microtransducer within a flaccid (800 ml) plastic bag. Gastric corpus-fundus pressure was recorded during distension of the bag with air (15 mils) over 30 s. Pressure indices were derived f r o m the areas under the pressure curves. Median (range) pressure indices were: healthy volunteers 12.7 (7.5-1 7.1) cmH,O, highly selective vagotomy 14.0 (9.8-15.9) cmH,O (n.s.), truncal vagotomy and drainage 14.5 (8.6-26.8) cmH,O (P = 0.04). All patients with diarrhoea had abnormally high pressure indices (P < 0.001). Pressure indices in all' other patient groups were within the normal range. W e conclude that gastric adaptive relaxation remains abnormal in patients with postvagotomy diarrhoea but not in those who are asymptomatic or who have other symptoms.
Impairment of gastric adaptive relaxation has been reported after both truncal ~ a g o t o m y ' -and ~ highly selective vagotomy6' and the degree of impairment is said to be similar after either ~ p e r a t i o n ' . ~ Studies . in both man and animals have been carried out after vagotomy, at l a p a r ~ t o m y ' - ~and ~ a few months later'.2*4*6.8*'2 . Few studies, however, have measured gastric adaptive relaxation more than 1 year after surgery. The impairment of gastric adaptive relaxation, in combination with other motility disturbances after vagotomy, is implicated in the pathophysiology of postvagotomy symptoms. Symptoms after truncal vagotomy and drainage tend to be worse than after highly selective vagotomy. This has been attributed to the more disturbed gastric emptying associated with the drainage procedure and denervation of the antropyloric segment and the small bowe1'3*'4. It is recognized, for example, that diarrhoea after highly selective vagotomy is uncommon, while after truncal vagotomy and drainage diarrhoea is frequently reported and may be a serious and persistent problem for a minority of patient^'^. The contribution of impaired gastric adaptive relaxation to the pathogenesis of postvagotomy disorders is both speculative and controversial. Postprandial epigastric fullness has been attributed to the failure of gastric adaptive relaxationis4. Early rapid gastric emptying is thought to be the key abnormality causing the dumping syndrome and may also contribute to the pathogenesis of postvagotomy diarrhoea. Early rapid gastric emptying is thought partly to be due to impaired gastric adaptive relaxation*. There are, therefore, clinically recognized links between impairment of gastric adaptive relaxation and the presence of symptoms after vagotomy. To date, there has been no study of gastric adaptive relaxation specifically comparing patients with and without symptoms after vagotomy. The aim of this study was to measure gastric adaptive relaxation in patients after highly selective vagotomy and truncal vagotomy and drainage in relation to their various symptoms more than 1 year after surgery. The study was formally approved by the Royal Liverpool Hospital
24
Ethical Committee in Novlember 1985. Written informed consent was given by all subjects studied.
Patients and methods Fifteen healthy volunteers were studied (13 men and two women). Their median age and body-weight were 30 (range 2 2 4 1 ) years and 70 (range 5G79) kg. All healthy volunteers 'were asymptomaticon no medication. None had undergone any previous upper gastrointestiiial surgery. There were 33 patients in the study. All had previously undergone vagotomy for intractable duodenal ulceration, confirmed by endoscopy or barium meal and subsequently at laparotomy, or carried out for complications of duodenal ulcer. Twenty-five of the patients (20 men and five women) had had trilncal vagotomy and drainage (18 pyloroplasty and seven gastroenterostomy) and eight patients, highly selective vagotomy (six men and two women). The median age and body-weight for patients in the truncal vagotomy and drainage group were 51 (range 35-67) years and 63 (range 44-92) kg. The median age and body-weight for patients in the highly selective vagotomy group were 35 (range 2 3 4 7 ) years and '77 (range 59-91) kg. The median time since operation was 4 years 6 months (range 1-24 years) for patients after truncal vagotomy and drainage and 2 (range 1-3) years for patients after highly selective vagotomy. Five patients after highly selective vagotomy and 12 patients after truncal vagotomy and drainage were asymptomatic. Symptoms in the remaining patients, summarized in Table 1 , had been present for over 1 year. Some symptomatic patients had more than one symptom. Table 2 shows the nature and si:verity of diarrhoea in patients with this symptom. No patient had or gave a history of recurrent ulceration following vagotomy. All symptomatic patients had undergone routine investigationof their symptoms. All patients studied were systematically classified in terms of presence or absence, nature and severity of symptoms before undergoing the studies described. The healthy volunteer and patient groups were studied in the same manner. The technique for measuring gastric adaptive relaxation was a modification of that used by Jahnberg et al.3 and identical to that recently reported by our groupi6. After an overnight fast, subjects swallowed, via the nose, a 10 Fr Ryle's tube with a collapsed plastic bag (800ml) sealed over the end, the side holes of the tube being within the bag. Within the lumen of the tube, positioned near the tip, was a
0007-1323/9l/OlOO24-04
(i? 1991 Butterworth-Helnemann Ltd
Gastric adaptive relaxation and symptoms after vagotomy: M . N. Hartley and C. R. Mackie
Table 1 Symptoms in patients after vagotomy
Highly selective vagotomy Truncal vagotomy and drainage
No. of patients
Symptoms
5 3 12 5 3 2 I
None Epigastric fullness None Epigastric fullness Bilious vomiting Vasomotor dumping Diarrhoea
median pressure index of 13.8 cmH,O (Figure 3). The median pressure index for the symptomatic patients after truncal vagotomy was 17.6 cmH,O and this was significantly higher ( P < 0.025) than both healthy volunteers and asymptomatic patients. Further analysis of symptoms in relation to pressure indices
....................................
Table 2 Characteristics of postvagotomy diarrhoea in seven patients after truncal vagotomy and drainage
180
In s u f f l a t i o n
Diarrhoea Continuous/episodic
Stool frequency (maximum per day)
Time I s )
Urgency Dynamic pressure index = Area under curve during insufflation
+
Continuous Continuous Continuous Continuous Episodic Episodic Episodic
After insufflation
+ +++ +++ +++ + ++
Continuous, loose stools every day; episodic, loose stools at intervals of days or weeks; urgency: +, urge to defaecate; + + , precipitate urge to defaecate; + +, incontinent of faeces
c m ~ Z ~
30 ’
Figure 1 Derivation of the dynamic pressure index from the pressure recording
30
+
pressure microtransducer (Gaeltec, Dingwall, Rosshire, UK) which was calibrated against a water manometer to provide a full scale deflection of 50 cm of water. The transducer was interfaced with an amplifier and pen recorder (Could Electronics Ltd., Coventry, UK) to record pressure changes against time. The bag was positioned manometrically in the corpus-fundus of the stomach. A baseline recording was obtained during a 15 min rest period. Air from a compressed air cylinder (British Oxygen Company, Guildford, UK) was delivered at a constant rate via an air flow meter and a safety valve to the Ryle’s tube. This constant flow of air was used to distend the hag at a rate of 15 ml/s for 30 s during which the pressure rise within the intragastric bag was recorded. The recording was continued at this end volume for a further 150 s. The air was then aspirated with a syringe and the volume noted. This procedure was repeated at 15 min intervals to provide four recordings for each subject, two each with subjects sitting and supine. For each recording a dynamic pressure index was derived from the area under the curve during distension. These measurements were ascribed units in cmH,O (Figure I). A single value was then derived for each subject by taking the mean of the four observations. Statistical analyses of pressure indices between groups were made using the Kruskal-Wallis and Mann-Whitney I/ tests. Analysis of variance was used to compare volumes of air aspirated between groups. The 1’ test was used to compare the types of drainage procedure.
25
za
1!
0
: I
0
l 0
1(
0
Results As previously reported” there was no identifiable trend of variation in successive recordings during individual studies. There was no detectable difference between sitting and supine recordings. All pressure indices in patients after highly selective vagotomy were within the range obtained for healthy volunteers (Figure 2). The median pressure index for these patients (14.0 cmH,O) was higher than that for the healthy volunteer group (12.7 cmH,O) but this difference was not statistically significant. Pressure indices in patients after truncal vagotomy and drainage were significantly higher than those in the healthy volunteer group (P=0.04). The median pressure index for patients after truncal vagotomy and drainage was 14.5 cmH,O (Figure 2). When these patients were divided into asymptomatic and symptomatic groups, all of the asymptomatic patients had pressure indices within the range for healthy volunteers with a
Br. J. Surg.. Vol. 78, No. 1, January 1991
I Controls
I Highly selective vagotorny
I T r u n c a l vagotorny a n d drainage
Figure 2 Scattergram of pressure indices in healthy volunteers and patienls after highly selective vagotomy and truncal vagotomy and drainage. Median values shown by bars. ( 0 ) Asymptomatic; (H) symptomatic
25
Gastric adaptive relaxation and symptoms after vagotomy: M. N. Hartley and C. R. Mackie
different (51 years (median; range 39-67 years) and 65 (44-92) kg respectively). There was no statistical correlation between the type of drainage procedure and the presence or absence of postvagotomy diarrhoea; four of 18 patients without diarrhoea and three of seven patients with diarrhoea had a gastroenterostomy.
30
25
Discussion
ON
I
-5 X
g .-C
+
2c
1:
a
. I . .. . .
-c
3
"1 Lo
L
CL
11
t
8
I
Control
.. . .. . .
8
8
-
. . -
I I I Asymptomatic Symptomatic T r u n c a l t r unca I truncal vagotorny vagotorny vagotorny and drainage a n d d r a in a g e a n d d r a in a g e ( d i a r r h o e a )
Figure 3 Scattergram of pressure indices in healthy volunteers and asymptomatic and symptomatic patients afer truncal vagotomy and drainage. Patients with postoagotomy diarrhoea are shown separately. Median values shown by bars. (e)Asymptomatic; (m) symptomatic
revealed that patients with epigastric fullness tended to have higher pressure indices in both vagotomy groups; three patients with epigastric fullness in the highly selective vagotomy group were within the top four pressure indices of their group (Figure 2 ) and four of the five patients with epigastric fullness after truncal vagotomy had pressure indices above the range for healthy volunteers. The latter patients also had postvagotomy diarrhoea. Patients with bilious vomiting and vasomotor dumping had pressure indices within the range for healthy volunteers. All seven patients with postvagotomy diarrhoea had pressure induces above the range for healthy volunteers. The median pressure index for the patients in this group was 20.3 cmH,O. This value was significantly higher (P
