Gut, 1978, 19, 358-366
Pancreatic enzymes in human duodenal juice-a comparison of responses in
secretin pancreozymin
and Lundh Borgstrom tests JOAN M. BRAGANZA, K. HERMAN, PAULINE HINE, G. KAY, AND G. I. SANDLE From the University Department of Gastroenterology, and the Department of Medical Physics, Manchester Royal Infirmary, Manchester
We have compared the concentrations and total outputs of trypsin, amylase, and lipase secreted into the duodenum during secretin pancreozymin and Lundh Borgstrom tests in order to assess the relative intensity of pancreatic stimulation provided by these means. The mean concentration of trypsin, and the mean 10 minute total output of trypsin, amylase, and lipase after the meal (though less than the peak post-prandial responses) were equal to the respective peak enzyme responses after intravenous injection of 2 Crick Harper Raper units per kg of secretin Boots; and after intravenous injection of 2 Crick Harper Raper units per kg of pancreozymin Boots.
SUMMARY
We are currently assessing the relative usefulness Harper Raper (CHR) units per kg of pancreozymin of the secretin pancreozymin test (Burton et al., Boots (batch numbers 5476, 5833) and duodenal 1960) and the Lundh Borgstr6m test (Lundh, 1962; juice was aspirated in 10 minute fractions for 30 Lundh and Borgstrom, 1962) in diagnosis and minutes after each injection. In the Lundh Borgstrom management of pancreatic disease. Preliminary test 500 ml of a standard meal (Borgstr6m et al., results in normal subjects enabled us to compare the 1957) was ingested over 10 minutes, and duodenal concentrations of pancreatic enzymes in duodenal juice was collected for four successive 30 minute juice after the physiological stimulus of a test periods. meal, and after intravenously administered secretagogues. A formal study was then planned to SUBSEQUENT STUDY measure total outputs of pancreatic enzymes In the subsequent study a quantitative secretin pancreozymin test was performed in three informed secreted during the two tests. volunteers using the double marker technique of Go et al. (1970) which involves, after simultaneous Methods intubation of stomach and duodenum, constant perfusion of the stomach at 1 5-2 ml per minute INITIAL STUDY The initial study was performed in 12 patients with isotonic saline containing a non-absorbable without pancreatic disease, and three normal marker, 14C labelled polyethylene glycol (New volunteers. Each subject received a secretin pancreo- England Nuclear Co), and constant perfusion of the zymin test (SP test), and on a separate day a Lundh duodenum at 1 5-2 ml per minute with isotonic saline Borgstrom test; the order of tests was randomised. (pH 6) containing another non-absorbable marker, Both tests were performed as designed by the polyethylene glycol (molecular weight 4000 BDH originators. In the SP test pancreatic secretion was Chemicals Ltd.). After an initial 40 minute equistimulated by intravenous injection of 2 Crick libration period duodenal contents, recovered by Harper Raper (CHR) units per kg of secretin means of an intermittent suction pump, were Boots (batch numbers 5360, 5664) followed 30 pooled into 10 minute fractions. Each volunteer minutes later by intravenous injection of 2 Crick was also given a quantitative Lundh Borgstrom test using a technique designed by Malagelada et al. (1976). The test meal which contained 2-5 ,tc Received for publication 9 December 1977 358
Pancreatic enzymes in human duodenal juice
14C labelled PEG was ingested over 10 minutes and subsequently at 10 minute intervals 50 ml of gastric contents were aspirated by syringe, mixed thoroughly, a 10 ml aliquot retained for analysis, and the remaining 50 ml returned to the stomach. At the end of 120 minutes the stomach was emptied and the volume of gastric contents recorded. Throughout the test the duodenum was perfused at 1 5-2 ml per minute with isotonic saline containing PEG (5 g/l) and duodenal contents were aspirated in 10 minute fractions. Chemical methods The pH of all gastric and duodenal samples was measured; concentration of trypsin was estimated by the pH stat method of Haverback et al. (1960) with p-toluene sulphonyl-L-argininemethyl ester (TAME) as substrate, amylase by Lagerlof's modification of the saccharogenic method of Norby (Lagerlof, 1942), lipase by a pH stat method with an olive oil emulsion as substrate (MarchisMouren et al., 1959), and bilirubin by the method of Powell (1944). Concentrations of PEG 4000 in gastric and duodenal samples were determined by the method of Hyden (1955) and concentrations of 14C by liquid scintillation counting. Ten duodenal samples obtained during SP and Lundh Borgstr6m tests were analysed in duplicate. The coefficients of variation in the determination of PEG, 14C counts, amylase, lipase and trypsin were 1 0, 3 0, 2-0, 2-9, and 5 8 respectively. In in vitro experiments 5 mg PEG 4000 per ml of duodenal juice was found not to inhibit amylolytic proteolytic or lipolytic activity.
Analytical methods The volume flowing past the sampling site in the distal duodenum was calculated from the rate of infusion of PEG saline into the proximal duodenum and the ratio of PEG concentration in the infusate and in aspirated samples (Go et al., 1970). Details of calculations to determine the total outputs of secreted pancreatic enzymes in studies employing gastric and duodenal markers have been provided by Go et al. (1970), and Malagelada et al. (1976). If it is assumed that complete mixing of duodenal marker and secreted pancreatic enzymes occurs, the volumes of duodenal juice regurgitating into the stomach in each 10 minute period of the SP test can be calculated by dividing the PEG 4000 content in successive gastric aspirates by the concentration of PEG 4000 in respective duodenal aspirates. We have not applied any systematic correction for reflux during the course of the Lundh Borgstrom test. Instead the volume of duodenal juice present in gastric contents aspirated
359 at the end of the test was calculated, and multiplied by the mean 10 minute concentration of each pancreatic enzyme in duodenal aspirates. This product was added to the cumulative two hour output of each enzyme in duodenal samples and the mean 10 minute total enzyme output was determined
(Table 5). We have used the log values of enzyme concentrations and outputs in this study, as the distribution of pancreatic enzyme output, which in a normal population is skew (Sarles et al., 1963; Ammann et al., 1968; Goldberg and Wormsley, 1970), can be normalised by logarithmic transformation (Sarles et al., 1963: Sun, 1963, Ammann et al., 1968). The concentrations and outputs of enzymes in response to meal and hormonal stimulation have been compared by paired t tests.
Results COMPARISON OF CONCENTRATIONS OF PANCREATIC ENZYMES IN SP AND LUNDH BORGSTROM TESTS
As the effect of bolus intravenous injections of secretin and pancreozymin on enzymes is short lived (Fig. 1) peak enzyme concentrations, which invariably occurred within the first 10 minutes, were taken to represent enzyme responses to hormonal stimulation. In the Lundh Borgstr6m test the peak concentration of each enzyme, which in an individual subject could occur 30, 60, 90, or 120 minutes after the meal (Table 1), was significantly higher than the mean concentration of the respective enzyme in pooled two hour collections of duodenal juice (Figs 2, 3, 4). Individual values for mean and S P TEST IZ
7.6 S. SE 4
76
6'
U
Ul
wU
z
5.
'E 4
E 4
x 3
3. 3.
C,
3.
I..
D
wJ 2VI
0%
.s1 2.
0%
cc
'I
''
- °'
o --
30
o
60 '
60
MINUTES
Fig. 1 Mean concentrations of trypsin ( x -x ), amylase (O --- 0), and lipase (A- .-A) in successive 10 minute collections of duodenal juice during SP tests in 15 patients.
Joan M. Braganza, K. Herman, Pauline Hine, G. Kay, and G. I. Sandle
360
Table. 1 Concentrations of trypsin, amylase, and lipase in four successive 30 minute collections of duodenal juice during Lundh Borgstrom test in 15 subjects N
=
Trypsin (IU/rm)
15
30'
M.P. P.M. C.P. J.R. H.W. I.S. M.W. W.B. M.M. A.B. E.C. T.M. G.S. R.K. J.H.
60' 43 0 166'2 34 5 22-2 88-0 54 5 44-7 70-2 81-7 54 7
41-5 54'7
36'0 31'2 102'5 33 0 28-2 58 7
104'0 13-0 28 5
37'5 116-5 22-0
119'5 44-0 29-0
77-5
31'5
31t5
Amylase (IU/ml) 90' 54 0 89 5
110'5 40-2 81-5
52'S 510 137'7 51'2
120'
30'
60'
90'
120'
30'
122'2 99-2
140-8 143-4
135-7 839'6
235'5 580 6
245 7 5529
84-7
112-6 99-8 143-3 122-8 76-8
61-4 69-1 151-0 135-6 230-4
2714 30 7 194-5
138-2 30 7
194-5 84'5
220'1
219 389 189 179 589 242 138 273 417 74 134 447 362 278 243
31'2
100-0 44-5 29-7 83-0 81-7
18-5
41-5
52'5
68'5
98 5
99.7 51-5 16-0
56-5 37 5
27'5
Lipase (IU/ml)
41'5
245'7 58 8 148'4 76-8
194'S 184'3
225'3
171'5
307-2
220-1 110.0 240 6
204'8
184-3
199-6 174-0 286-7 665 6 199'6 117'7
716'8 168-9 117'7
112'6 148'4 122 8
92'1
99-8
645.1
450'5
512-0
122'8
153-6
174-0
60' 219
1674 184 173 405 297 263 323 224 376 204 447 180 725 234
90'
120'
395 944 794 148 440 324 235 263 162 157 414 262 445 570 193
515 664 442 99 347 262 143 303 245 442 300 382 440 523 266
Peak values are in italic.
peak concentrations of enzymes after the meal, and peak concentrations after secretin, and after pancreozymin are given in Table 2. Groups were compared by paired t tests after logarithmic transformation of data and the results are shown in Figs 2, 3, 4. For trypsin mean concentration after meal equalled peak concentration after secretin, and after pancreozymin, but peak response after meal significantly exceeded the response to either hormone (Fig. 2, Table 2); for amylase meal mean and peak responses were not significantly different from responses to hormones (Fig. 3, Table 2), while for lipase both meal mean and peak responses were significantly less than the values after secretin or pancreozymin (Fig. 4, Table 2). MEAL
20
MEAN
SECRETIN PEAK
PEAK
MEAL MEAN
SECRETIN PEAK
PEAK
PANCREOZYMIN PEAK
100
~o
0.6 < p
Pancreatic enzymes in human duodenal juice-a comparison of responses in
secretin pancreozymin
and Lundh Borgstrom tests JOAN M. BRAGANZA, K. HERMAN, PAULINE HINE, G. KAY, AND G. I. SANDLE From the University Department of Gastroenterology, and the Department of Medical Physics, Manchester Royal Infirmary, Manchester
We have compared the concentrations and total outputs of trypsin, amylase, and lipase secreted into the duodenum during secretin pancreozymin and Lundh Borgstrom tests in order to assess the relative intensity of pancreatic stimulation provided by these means. The mean concentration of trypsin, and the mean 10 minute total output of trypsin, amylase, and lipase after the meal (though less than the peak post-prandial responses) were equal to the respective peak enzyme responses after intravenous injection of 2 Crick Harper Raper units per kg of secretin Boots; and after intravenous injection of 2 Crick Harper Raper units per kg of pancreozymin Boots.
SUMMARY
We are currently assessing the relative usefulness Harper Raper (CHR) units per kg of pancreozymin of the secretin pancreozymin test (Burton et al., Boots (batch numbers 5476, 5833) and duodenal 1960) and the Lundh Borgstr6m test (Lundh, 1962; juice was aspirated in 10 minute fractions for 30 Lundh and Borgstrom, 1962) in diagnosis and minutes after each injection. In the Lundh Borgstrom management of pancreatic disease. Preliminary test 500 ml of a standard meal (Borgstr6m et al., results in normal subjects enabled us to compare the 1957) was ingested over 10 minutes, and duodenal concentrations of pancreatic enzymes in duodenal juice was collected for four successive 30 minute juice after the physiological stimulus of a test periods. meal, and after intravenously administered secretagogues. A formal study was then planned to SUBSEQUENT STUDY measure total outputs of pancreatic enzymes In the subsequent study a quantitative secretin pancreozymin test was performed in three informed secreted during the two tests. volunteers using the double marker technique of Go et al. (1970) which involves, after simultaneous Methods intubation of stomach and duodenum, constant perfusion of the stomach at 1 5-2 ml per minute INITIAL STUDY The initial study was performed in 12 patients with isotonic saline containing a non-absorbable without pancreatic disease, and three normal marker, 14C labelled polyethylene glycol (New volunteers. Each subject received a secretin pancreo- England Nuclear Co), and constant perfusion of the zymin test (SP test), and on a separate day a Lundh duodenum at 1 5-2 ml per minute with isotonic saline Borgstrom test; the order of tests was randomised. (pH 6) containing another non-absorbable marker, Both tests were performed as designed by the polyethylene glycol (molecular weight 4000 BDH originators. In the SP test pancreatic secretion was Chemicals Ltd.). After an initial 40 minute equistimulated by intravenous injection of 2 Crick libration period duodenal contents, recovered by Harper Raper (CHR) units per kg of secretin means of an intermittent suction pump, were Boots (batch numbers 5360, 5664) followed 30 pooled into 10 minute fractions. Each volunteer minutes later by intravenous injection of 2 Crick was also given a quantitative Lundh Borgstrom test using a technique designed by Malagelada et al. (1976). The test meal which contained 2-5 ,tc Received for publication 9 December 1977 358
Pancreatic enzymes in human duodenal juice
14C labelled PEG was ingested over 10 minutes and subsequently at 10 minute intervals 50 ml of gastric contents were aspirated by syringe, mixed thoroughly, a 10 ml aliquot retained for analysis, and the remaining 50 ml returned to the stomach. At the end of 120 minutes the stomach was emptied and the volume of gastric contents recorded. Throughout the test the duodenum was perfused at 1 5-2 ml per minute with isotonic saline containing PEG (5 g/l) and duodenal contents were aspirated in 10 minute fractions. Chemical methods The pH of all gastric and duodenal samples was measured; concentration of trypsin was estimated by the pH stat method of Haverback et al. (1960) with p-toluene sulphonyl-L-argininemethyl ester (TAME) as substrate, amylase by Lagerlof's modification of the saccharogenic method of Norby (Lagerlof, 1942), lipase by a pH stat method with an olive oil emulsion as substrate (MarchisMouren et al., 1959), and bilirubin by the method of Powell (1944). Concentrations of PEG 4000 in gastric and duodenal samples were determined by the method of Hyden (1955) and concentrations of 14C by liquid scintillation counting. Ten duodenal samples obtained during SP and Lundh Borgstr6m tests were analysed in duplicate. The coefficients of variation in the determination of PEG, 14C counts, amylase, lipase and trypsin were 1 0, 3 0, 2-0, 2-9, and 5 8 respectively. In in vitro experiments 5 mg PEG 4000 per ml of duodenal juice was found not to inhibit amylolytic proteolytic or lipolytic activity.
Analytical methods The volume flowing past the sampling site in the distal duodenum was calculated from the rate of infusion of PEG saline into the proximal duodenum and the ratio of PEG concentration in the infusate and in aspirated samples (Go et al., 1970). Details of calculations to determine the total outputs of secreted pancreatic enzymes in studies employing gastric and duodenal markers have been provided by Go et al. (1970), and Malagelada et al. (1976). If it is assumed that complete mixing of duodenal marker and secreted pancreatic enzymes occurs, the volumes of duodenal juice regurgitating into the stomach in each 10 minute period of the SP test can be calculated by dividing the PEG 4000 content in successive gastric aspirates by the concentration of PEG 4000 in respective duodenal aspirates. We have not applied any systematic correction for reflux during the course of the Lundh Borgstrom test. Instead the volume of duodenal juice present in gastric contents aspirated
359 at the end of the test was calculated, and multiplied by the mean 10 minute concentration of each pancreatic enzyme in duodenal aspirates. This product was added to the cumulative two hour output of each enzyme in duodenal samples and the mean 10 minute total enzyme output was determined
(Table 5). We have used the log values of enzyme concentrations and outputs in this study, as the distribution of pancreatic enzyme output, which in a normal population is skew (Sarles et al., 1963; Ammann et al., 1968; Goldberg and Wormsley, 1970), can be normalised by logarithmic transformation (Sarles et al., 1963: Sun, 1963, Ammann et al., 1968). The concentrations and outputs of enzymes in response to meal and hormonal stimulation have been compared by paired t tests.
Results COMPARISON OF CONCENTRATIONS OF PANCREATIC ENZYMES IN SP AND LUNDH BORGSTROM TESTS
As the effect of bolus intravenous injections of secretin and pancreozymin on enzymes is short lived (Fig. 1) peak enzyme concentrations, which invariably occurred within the first 10 minutes, were taken to represent enzyme responses to hormonal stimulation. In the Lundh Borgstr6m test the peak concentration of each enzyme, which in an individual subject could occur 30, 60, 90, or 120 minutes after the meal (Table 1), was significantly higher than the mean concentration of the respective enzyme in pooled two hour collections of duodenal juice (Figs 2, 3, 4). Individual values for mean and S P TEST IZ
7.6 S. SE 4
76
6'
U
Ul
wU
z
5.
'E 4
E 4
x 3
3. 3.
C,
3.
I..
D
wJ 2VI
0%
.s1 2.
0%
cc
'I
''
- °'
o --
30
o
60 '
60
MINUTES
Fig. 1 Mean concentrations of trypsin ( x -x ), amylase (O --- 0), and lipase (A- .-A) in successive 10 minute collections of duodenal juice during SP tests in 15 patients.
Joan M. Braganza, K. Herman, Pauline Hine, G. Kay, and G. I. Sandle
360
Table. 1 Concentrations of trypsin, amylase, and lipase in four successive 30 minute collections of duodenal juice during Lundh Borgstrom test in 15 subjects N
=
Trypsin (IU/rm)
15
30'
M.P. P.M. C.P. J.R. H.W. I.S. M.W. W.B. M.M. A.B. E.C. T.M. G.S. R.K. J.H.
60' 43 0 166'2 34 5 22-2 88-0 54 5 44-7 70-2 81-7 54 7
41-5 54'7
36'0 31'2 102'5 33 0 28-2 58 7
104'0 13-0 28 5
37'5 116-5 22-0
119'5 44-0 29-0
77-5
31'5
31t5
Amylase (IU/ml) 90' 54 0 89 5
110'5 40-2 81-5
52'S 510 137'7 51'2
120'
30'
60'
90'
120'
30'
122'2 99-2
140-8 143-4
135-7 839'6
235'5 580 6
245 7 5529
84-7
112-6 99-8 143-3 122-8 76-8
61-4 69-1 151-0 135-6 230-4
2714 30 7 194-5
138-2 30 7
194-5 84'5
220'1
219 389 189 179 589 242 138 273 417 74 134 447 362 278 243
31'2
100-0 44-5 29-7 83-0 81-7
18-5
41-5
52'5
68'5
98 5
99.7 51-5 16-0
56-5 37 5
27'5
Lipase (IU/ml)
41'5
245'7 58 8 148'4 76-8
194'S 184'3
225'3
171'5
307-2
220-1 110.0 240 6
204'8
184-3
199-6 174-0 286-7 665 6 199'6 117'7
716'8 168-9 117'7
112'6 148'4 122 8
92'1
99-8
645.1
450'5
512-0
122'8
153-6
174-0
60' 219
1674 184 173 405 297 263 323 224 376 204 447 180 725 234
90'
120'
395 944 794 148 440 324 235 263 162 157 414 262 445 570 193
515 664 442 99 347 262 143 303 245 442 300 382 440 523 266
Peak values are in italic.
peak concentrations of enzymes after the meal, and peak concentrations after secretin, and after pancreozymin are given in Table 2. Groups were compared by paired t tests after logarithmic transformation of data and the results are shown in Figs 2, 3, 4. For trypsin mean concentration after meal equalled peak concentration after secretin, and after pancreozymin, but peak response after meal significantly exceeded the response to either hormone (Fig. 2, Table 2); for amylase meal mean and peak responses were not significantly different from responses to hormones (Fig. 3, Table 2), while for lipase both meal mean and peak responses were significantly less than the values after secretin or pancreozymin (Fig. 4, Table 2). MEAL
20
MEAN
SECRETIN PEAK
PEAK
MEAL MEAN
SECRETIN PEAK
PEAK
PANCREOZYMIN PEAK
100
~o
0.6 < p
