- Email: [email protected]
Pylorus-preservation decreases the extent of atrophy of the remnant pancreas after pancreatoduodenectomy
rl PB 1999
Vol ume I, Numbe r 2, 65- 70
Pylorus-preservation decreases the extent of atrophy of the remnant pancreas after pancreatoduodenectomy s-w Kiml, KH Kiml,JK Han 2 andYH Departments
Park l
of I Surgery and 2Radiology, Seoul National University College of Medicine, Seoul, Korea
Background Pylorus-preserving pancreatoduodenectomy (PPPD) pre-
or PPPD performed by the same surgeon an d w ho had survived ;:0: I year without tumour recurrence. For vo l-
serves the secretion of gastrointestinal (GI) hormones
umetry, thin-section spiral CT with 3D display was per-
from the distal stomach and duodenum , whereas after pan-
formed; hormone release was meas ured by radioim uno-
creatoduodenectomy (PD) they are no longer secreted. It has been suggested that some GI hormones exert a troph-
assay using antibodies.
Results
ic effect on the pancreas, although this effect has not been
After PPPD, pancreatic vo lum e and gastri n and CCK
documented in man. It was postulated that the ablation of
release were significantly greater than after PD; there was
GI hormones, such as gastrin and cholecystokinin (CCK),
a significant co r relation between pancreatic volume and
by PD is an important cause of postoperative pancreatic
stimulated gastr in re lease . Pancreatic volume was not
atrophy and, since PPPD preserves the secretion of these
related to other clinical factors , such as type of re con-
hormones, it would be more effective than PD for the
struction , age , post operative interval, or nutritional status.
maintenance of postoperative pancreatic function. Our study aimed to determine whether pylorus preservation
Discussion The volume of distal remnant pancreas is greater after
after PD affects the volume of the remnant pancreas in
PPPD than after PD and this diffe re nce may reflect preser-
long-term survivors and whether any such effect is related
vation (by PPPD) of GI hormones that exert a trophic
to the continued secretion of GI hormones following preservation of the pylorus.
Methods We measured postoperative pancreatic volume and the
effect on the pancreas.
Keywords pylorus preservation, re mnant pan creas atrophy, pancreatoduodenectomy.
release of gastrin and CCK in patients who underwent PD
Introduction
Whipple operation (PD) for most periampullary cancers.
Postoperative diarrhoea, steatorrhea and poor n utritional status are not infrequently seen afte r pancreatoduodenecto-
The various advantages of PPPD over PD, including postoperative n utrition al status, have been well documented
my (PD) and appear to be the result of partial gastrectomy and pancreatic exocrine insufficiency [1] . Atrophy of the re mnant distal pancreas is also frequen t and might be caused by ph ys iological ch ange in normal digestive func-
and relate to th e preservation of gastric capacity [3,4]. Several investigators h ave sh own that PPPD preserves the secretion of gastrointestinal (GI) hormones, such as gastrin, secretin and cholecystokinin (CCK) from the dis-
tion, or poor pancreatic drainage through the site of the pancreatoenteric anastomosis.
tal stomach and proximal duodenum; after PD their secretion almost completely stops [5- 8]. These differences could
Pylorus-preserving pan creatoduoden ectomy (PPPD) was first used in cases of chronic pancreatitis [2]. More
be associated with better nutrition after PPPD. It h as been suggested that some GI hormones have a trophic effect on the mucosa of the GI tract and pancreas
recently, PPPD h as tended to replace the con vention al Correspondence to: SW Kim, Deportment of Surgery, Seoul Notional University College of Medicine, 28 Yongon-Dong, Chongro-Ku, Seoul I 10-744, Korea
© I999 Isis Medical Media Ltd.
65
SW Kim et of. [9] and, in animal models, this physiological effect is well established [9-14]. Hypertrophic gastric mucosa in gastrinoma patients [15] provides good evidence of the trophic effect of GI hormones in humans. However, since it is almost impossible to establish a human model, the existence of this trophic effect has not been established . We postulated that the ablation of GI hormones, including gastrin and CCK, during PD is one of the important causes of postoperative pancreatic atrophy. Unlike PD, PPPD to some extent preserves these hormones, and so postoperative atrophy could be less marked. Better nutrition after PPPD would thus be due to the maintenance of pancreatic volume, as well as preservation of the pylorus.
cm of distal jejunum were anastomosed to the stomach or duodenum [16].
Pancreas volumetry Thin-section spiral CT was performed on the remnant distal pancreas; reconstructed images were used for 3D display and volumetric measurement. After localisation of the pancreas by conventional CT, spiral CT was performed on the region of the pancreas 30-65 s after the injection of i.v. contrast material at a speed of 3 cm 3 S- 1 and images were reconstructed at 5 mm intervals. The reconstructed CT image data were sent to a 3D graphic work-station (Allegro, ISG Technologies, Missisauga, C anada). On the work-station, a threshold was
To determine whether there is evidence for this hypothesis, postoperative pancreatic volume and stimulat-
set to include the area of the pancreas. Soft tissue areas other than the pancreas that were included by the thresh-
ed GI hormone release were measured in patients who underwent either PD or PPPD for periampullary cancer.
old were manually removed from each image by the abdominal radiologist. After the editing, a 3D object was generated and the volume was calculated by the work-station. To minimise the possible variation caused by the threshold setting, the threshold value was set to be constant in all
Methods Patients Patients who underwent PPPD (n
=
16) or PD (n
=
12) for
periampullary cancer performed by the same surgeon and who survived 2 1 year without evidence of recurrence were enrolled in this study. PPPD was the operation of choice; PD was performed if the lesion was of pancreatic origin, if the patient had gastric lesions, such as peptic ulcer or gastritis (either at the time of diagnosis or based on the medical history), or if the blood supply of the duodenum looked insufficient during PPPD. Patients with preoperative pancreatic atrophy were excluded.
Surgical techniques During PD or PPPD the pancreas was cut along the midline of the portal and superior mesenteric vein. Either pancreatogastrostomy (PG, n = 15) or pancreatojejunostomy (PJ, n = 13) was used for reconstruction. The pancreas was anastomosed and invaginated into the antrum or the low gastric body (posterior wall) for PG and into the antimesenteric side of the jejunum for PJ. Anastomosis was performed using a two-layer interrupted suture with 4-0 black silk; a polyethylene tube about 3 cm long was inserted into the pancreatic duct as a stent. The extent of gastric resection during PD was hemigastrectomy, which would not affect the choice of PG. The proximal jejunal end was closed and an end-to-side hepaticojejunostomy was performed; 40-50
66
patients, between 14- 175 Hounsfield units. If preoperative spiral CT was available, this too was measured, as was the pancreatic volume distal to the mid-line (portal and superior mesenteric vein).
Gastrin and CCK assay After patients fasten for at least 8 h, a blood sample was taken before and at 15, 30, 45 and 60 min after ingestion of an amino acid-enriched fluid diet (EnsureR 1 can; 200 ml, 400 k cal). Samples were collected in Trasylol and EDTA-treated tubes. Plasma was taken after centrifuge and was stored in the freezer until hormonal assay. Radioimmunoassays, using antiCCK antibody (OAL-656, Ostuka Assay Laboratory, Japan) and a gastrin kit (DPC, USA) were performed [17].
Evaluation of nutritional status Body weight change and current body mass index (BMI) were compared between PD and PPPD groups. The extent of recovery to the body weight on admission was stratified as 2 100%,95- 100% and < 95%. Body weight heighC 1 (kg m- 2 ) was used as BMI.
Statistical evaluation The results are expressed as mean ± SE and an independent t test was used to evaluate statistical significance. p < 0.05 was considered significant.
PPPD prevents pancreatic atrophy
Results The patients' clinical and nutritional profile, according to the type of operation, is shown in Tables 1 and 2. The nutritional status with respect to weight recovery and BMI was better in the PPPD group. However, these differences were only of the p = 0.07 and 0.06 level. The mean volume of the remnant distal pancreas after PPPD (21 S4S±2117 mm3 ) was greater than that after standard PD (13 Sl1 ±1219 mm 3 ) (p < 0.05). Figure 1 sh ows the mean volume of the remnant distal pancreas per body weight in the two groups of patients. Pancreatic volume per body weight after PPPD (388.0±39.9 mm3 kg- I) was greater than that after PD (263.3±27.3 mm3 kg- I) (p < 0.05) . Preoperative volumetry of the pancreas distal to the mid-line of the superior mesenteric-portal vein could be performed in 12 patients. There was no difference in volume per body weight between the PPPD group (n = 7; 771.4 ±122 .S mm 3 kg- I) and the PD group (n = 5; 727.S ±20S.6 mm3 kg- I). The extent of volume decrease, however, was much less in the PPPD group (42.9%) than in the PD group (69.4%) (p = 0.024) (Figure 2).
Basal and stimulated gastrin release in the PPPD group and standard PD group is shown in Figure 3. In the PPPD group, plasma gastrin increased from S0.3±6.7 pg ml- I to 68.6 ±6.0 pg ml- 1 at 15 min after stimulation, wh ereas in the standard PD group there was no stimulated gastrin release and throughout the time course gastrin levels were significantly lower than in the PPPD group. There was a significant correlation between the gastrin level at 15 min after stimulation and the volume of remnant pancreas (Pearson correlation coefficient, p < 0.05) (Figure 4). Basal CCK level was 3.6±1.3 pg mL- I in the standard PD group and 4.8±1.1 pg ml- 1 in the PPD group. 30 min after stimulation, peak CCK response was 7.7 ±2.1 pg mL- 1
Cfl
1 3 800 mm kg-
CIl Q)
U c 700 CIl Q.
600
C
CIl
c
500
~
400
E
• •
v----1~
·
Q)
-£ 300 '0 200 Q)
E 100 ::J
~
Table I. Patient details
= 12)
PO (n
PPPO (n
= 16)
•• • • •
•
0
PO (n=12)
PPPO (n = 16)
Figure I. Volume of the pancreas after each type of pancreatoduodeaomy. =0.017.
p
Mean age (years) Male : Female Disease entity Ampulla of Vater cancer Common bile duct cancer Pancreatic cancer Duodenal cancer PJfPG anastomosis Postoperative month (mean)
59.7 2:I
61.7 I: I
2 7 2
II
70
5/7 23.2 (12-46)
5 0 0 8/8 32.9 (13-65)
60 Q)
~
.>~ _ 0
'§ --; 50 Q) CIl Q.Q)
g .....
g40
CIl -2-Q. Q)
Q)
.,E; ~30 CIlIDO
.8 E20
Table 2. Body weight recovery after operation Body weight recover
PO
(n %ratio to body weight on admission ~ 100% 95-100% < 95% BMI (kg m-2)
= 12)
PPPO (n = 16)
Cfl ::J 0_
P
Cl... 0
> 10
o
PO
Preoperative volume : 2 (16.7) 7 (43.8) I (8.3) 3 (18.8) 9 (75.0) 6 (37.5) 20.5 ± 0.9 22.7 ± 0.7
727.5+205.6 mm3 kg-1 771.4+122 .5 mm3 kg-1 n=7 0.074 0.064
n=7
Figure 2. The volume change of distal pancreas after each type of pancreatoduodeaomy. Preoperative volume includes the pancreas distal to the midline of the portal and superior mesenteric vein. p = 0.024.
67
SW Kim et 01.
~
E
80
18
70
16
60
14
OJ
'I
50 E:: c
enC\l
40
PPPD (n = 16)
OJ
C\l
30
C\l
20
E (fJ 0:::
12
--l
.;::
i-
~.
0..
~ 0 0
PD(n=12)
8
C\l
6
C\l
4
E (fJ
.....
10
E 10 OJ
0::: 2
0
15
0
30
45
60
0
Min
0
15
30
45
60
Min Figure 3. Secretion
of gastrin after each type of pancreatoduodectomy.
p < 0.005.
Figure 5. Patterns of CCK secretion after each type of pancreatoduodectomy. The difference in stimulated CCK release was not statistically significant.
c 120 (pg mL- 1) 0
~
•
::l E 100
.~
CD
80
c
60
• •
~
'E
who have undergone PPPD than in th ose who have undergone conventional Po. Since postoperative atrophy of the remnant pancreas is not exceptional, we interpre t this observation as showing that there is less atrophy after PPPD
• •
than after PD. Preoperative volumetry was performed in only 12 of 26 patients. The CT scans of many patients
L!)
Qi
40
•
>
.S!? c
20
"5 (fJ C\l
C)
00
100
...'
• • 200
300
Pancreatic
• 400
• 500
600
700
800
volume(mm 3 kg- 1)
Figure 4. Correlation between gastrin level 15 min after stimulation and volume of the remnant pancreas. There was significant correlation between these two factors (Pearson correlation coefficient =0.4815, P = 0.013).
in the standard PO group and 13.3±3.9 pg mL- i in the PPPD group. The stimulatory response of CCK was higher in the PPPD group than in the PO group, though the difference was not statistically significant (Figure 5). There was no correlation between CCK and pancreatic volume. Other factors that might have affected postoperative pancreative volume, namely type of reconstruction (PG versus PJ), age, postoperative interval and nutritional status (BM!) were investigated (Table 3). None correlated with pancreatic volume.
referred to us were no longer available, so their preoperative pancreatic volume could not be accurately checked. In this small group of 12 patients, however, there was a significant difference in postoperative volume of the pancreas, even though their preoperative volumes were similar. Better nutritional status after PPPD, as indicated by weight recovery and other parameters, h as been reported [1,2]. In this series, however, nutritional status between PPPD and PO were not significantly different although a trend in favour of PPPD was apparent. Nor was pancreatic volume per weight or body surface area significantly different; thus, nutritional factors do not appear to h ave any effect. Mainz and colleagues [10] were the fi rst to show that G I hormones , including CCK, increased pancreatic DN A, thus stimulating the growth of this organ. N umerous studies [9-12] have documented that gastrin and CCK or its an alogue stimulate the growth of the exocrine pancreas in
Discussion
different animal species, but this effect has not yet been demonstrated in man. Since long-term manipulation with exogenous or endogenous GI hormones is almost impossi-
This study has demonstrated, for the first time, rhat rhe \-01ume of the distal remnant pancreas is greater in patients
ble in men , suitable experiments would be very difficult to conduct.
68
PPPD prevents pancreatic atrophy
Table 3. Pan creative volume and clinical factors
Type of operation
PPPD PD
Pancreatic anastomosis
PJ PG :S;60
Age (years) Sex Postoperative adjuvant treatment Postoperative interval (months)
>60 Male Female Yes No :s;24 >24
n patients
Pancreas volume (mm 3 kg~') Mean±SE
16 12 13 15 13 15 16 12 II 17 12 16
388.0±39.9 263.3±27.3 330.7±38.s 334.1 ±40.3 386.2±47.8 282.9±24.4 3Is.2±3s.6 3s4.3±44.4 283.3±26.3 366.6±41.6 331.7±SO.S 333.3±30.9
p 0.017 NS 0.070 NS 0.141 NS
NS = not significant. SE = standard error.
After PPPD, many patients maintain their normal GI hormone response to stimuli, whereas after Whipple operation the secretion of hormones, such as gastrin, CCK and secretin, is negligible [5-8]. GI hormones might thus exert a trophic effect on the human pancreas. This effect is a possible explanation for the increased pancreatic volume found after PPPD.
Late symptoms of gastrointestinal disturbance, including diarrhoea, are usually not serious. It is very difficult to differentiate the causes of diarrhoea, since dumping syndrome [4]' pancreatic exocrine insufficiency, or dissection of the nerve plexus around the celiac and superior mesenteric artery give rise to similar symptoms. In our series of
Such a model might demonstrate the trophic effect of GI hormones in man. The results of this study suggest only a possible relationship. We showed that basal and stimulated gastrin release was greater in the PPPD group of patients, as previous reports have demonstrated [6-8]. Interestingly, there was a sig-
patients, only a few complained of intermittent diarrhoea, which was not clinically important and did not differ
nificant correlation between stimulated gastrin release and the volume of remnant pancreas, which suggests that the
well known, but whether pancreatic regeneration occurs after partial resection is uncertain. It has been demonstrated in an animal model that partial pancreatectomy stimulates remnant pancreatic growth, and that increased
extent of hormonal preservation may be related to the preservation of function and the volume of the target organ. Mean secretory CCK response was higher after PPPD, though CCK release was not seen in all cases. Whether
according to the type of resection or whether reconstruction was by PJ or PG. Regeneration of the liver after partial liver resection is
endogenous CCK enhances the regenerative effect [13,14]. Whether this same effect occurs in man remains uncertain.
CCK release is preserved depends, in part, on the length and function of the preserved first portion of the duodenum, which can vary from patient to patient. CCK can also
It may be assumed that distal pancreatectomy stimulates growth of the pancreatic head, whereas resection of the head, together with duodenum, might not stimulate growth
be released from the proximal jejunum and this fact can explain its release after PD. Pancreatic atrophy is a chronic postoperative change
of the distal pancreas because of the ablation of CCK release.
which can differ according to the interval between operation and the measurement of pancreatic volume. After
Underlying pancreatic pathology may affect the rate of proliferation of the remnant pancreas, though it is not clear
each operation type, however, the interval was similar and there was no correlation between length of interval and
whether chronic inflammation in the pancreas affects the extent of atrophy. In our series of patients, moderate-tosevere chronic inflammation in the margin of the resected
pancreatic volume.
pancreas was not detected.
69
SW Kim et 01. After pylorus preservation, originally performed to prevent the postgastrectomy syndrome, nutrition is more satisfactory [18] and this may also explain why there is less atrophy after PPPD. In this study, however, there was no differ-
6
7
ence in nutritional status and the absolute volume of the remnant pancreas did not correlate with nutritional status, BMI or body weight. In conclusion, the greater volume of the remnant pancreas after PPPD rather than PD may be due to the preservation of GI hormones, such as gastrin and CCK, that exert a trophic effect on the pancreas. As suggested in our previous report [16] the stimulatory effect of GI hormones on growth may affect anastomotic wound healing. Pyloric preservation results in the continuing secretion of GI hormones, thereby reducing anastomotic leakage after pancreatoduodenectomy.
Acknowledgement
8
Surg 1991;214:56-60. 9
10 11
12
Supported by grant number 95-010 from Seoul National University Hospital.
References
2
3
4
5
70
Traverso LW, Longmire WP. Preservation of the pylorus in pancreaticoduodenectomy. A follow-up evaluation. Ann Surg 1980;192:306-10. Traverso LW, Longmire WP. Preservation of the pylorus during pancreaticoduodenectomy. Surg Gynecol Obstet 1978; 146:959-62. Morel P, Mathey P, Corboud H et al. Pylorus-preserving duodenopancreatectomy: long-term complications and comparison with the Whipple procedure. World] Surg 1990;14:642-7. Braasch JW, Rossi RL, Watkins E Jr, Danniel DJ. Pyloric- and gastric-preserving pancreatic resection: experience with 87 patients. Ann Surg 1986;204:411-18. Pearlman NW, Stiegman GV, Ahnen DJ et al. Acid and gastrin levels following pyloric-preserving pancreaticoduodenectomy. Arch Surg 1986;121:661-4.
Saro T, Imamura M, Matsuno S et al. Gastric acid secretion and gut hormone release in patients undergoing pancreaticoduodenectomy. Surgery 1986;99:728-34. Takada T, Yasuda H, Shikaka J et al. Postprandial plasma gastrin and secretin concentrations after a pancreatoduodenectomy: a comparison between a pylorus-preserving pancreatoduodenecromy and the Whipple procedure. Ann Surg 1989;210:47-51. Tangoku A, Nishikawa M, Adachi A, Suzuki T. Plasma gastrin and cholecystokinin response after pylorus preserving pancreatoduodenectomy with Billroth-I type of reconstruction. Ann
13
Johnson LR. Regulation of gastrointestinal growth. In: Johnson DR editor. Physiology of the gastrointestinal tract. 2nd ed. New York: Raven Press, 1987:301-3. Maim DL, Black 0, Webster PD. Hormonal control of pancreatic growth.] Clin Invest 1973;52:2300-4. Dembinski AB, Johnson LR. Stimulation of pancreatic growth by secretin, caerulein and pentagastrin. Endocrinology 1980;105:323-8. Morriset J, Chamberland S, Gilbert L et al. Study of pancreatic DNA synthesis in vivo and in vitro following caerulein treatment in vivo. Biomed Res 1982;3:151-8. Johnson LR, Dudrick SJ, Guthrie PD. Stimulation of pancreatic growth by intraduodenal amino acids and HC!. Am ]
PhysioI1980;239:G400-5. 14 Lehv M, Fitzgerald PJ . Pancreatic acinar cell regeneration. IV. Regeneration after surgical resection. Am ] Pathol 1986;53:513-35. 15 Neuburger P, Lewin M, de Recherche C et al. Parietal and chief cell populations in four cases of Zollinger-Ellison syndrome. Gastroenterology 1972;63:937-42 16 Kim SW, Youk EG, Park YH. Comparison of pancreatogastrostomy and pancreatojejunostomy after pancreatoduodenectomy performed by one surgeon. World] Surg 1997;21:640-2. 17 Hashimura E, Shimizu F, Nishino T et al. Production of rabbit antibody specific for aminoterminal residues of cholecystokinin octapeptide (CCK-8) by selective suppression of crossreactive antibody response. ] Immunol Method 1982;5:375-87. 18 Linehan IP, Russell RCG, Hobsley M. The dumping syndrome after pancreaticojejunostomy. Surg Gynecol Obstet 1988;167: 114-18.
Vol ume I, Numbe r 2, 65- 70
Pylorus-preservation decreases the extent of atrophy of the remnant pancreas after pancreatoduodenectomy s-w Kiml, KH Kiml,JK Han 2 andYH Departments
Park l
of I Surgery and 2Radiology, Seoul National University College of Medicine, Seoul, Korea
Background Pylorus-preserving pancreatoduodenectomy (PPPD) pre-
or PPPD performed by the same surgeon an d w ho had survived ;:0: I year without tumour recurrence. For vo l-
serves the secretion of gastrointestinal (GI) hormones
umetry, thin-section spiral CT with 3D display was per-
from the distal stomach and duodenum , whereas after pan-
formed; hormone release was meas ured by radioim uno-
creatoduodenectomy (PD) they are no longer secreted. It has been suggested that some GI hormones exert a troph-
assay using antibodies.
Results
ic effect on the pancreas, although this effect has not been
After PPPD, pancreatic vo lum e and gastri n and CCK
documented in man. It was postulated that the ablation of
release were significantly greater than after PD; there was
GI hormones, such as gastrin and cholecystokinin (CCK),
a significant co r relation between pancreatic volume and
by PD is an important cause of postoperative pancreatic
stimulated gastr in re lease . Pancreatic volume was not
atrophy and, since PPPD preserves the secretion of these
related to other clinical factors , such as type of re con-
hormones, it would be more effective than PD for the
struction , age , post operative interval, or nutritional status.
maintenance of postoperative pancreatic function. Our study aimed to determine whether pylorus preservation
Discussion The volume of distal remnant pancreas is greater after
after PD affects the volume of the remnant pancreas in
PPPD than after PD and this diffe re nce may reflect preser-
long-term survivors and whether any such effect is related
vation (by PPPD) of GI hormones that exert a trophic
to the continued secretion of GI hormones following preservation of the pylorus.
Methods We measured postoperative pancreatic volume and the
effect on the pancreas.
Keywords pylorus preservation, re mnant pan creas atrophy, pancreatoduodenectomy.
release of gastrin and CCK in patients who underwent PD
Introduction
Whipple operation (PD) for most periampullary cancers.
Postoperative diarrhoea, steatorrhea and poor n utritional status are not infrequently seen afte r pancreatoduodenecto-
The various advantages of PPPD over PD, including postoperative n utrition al status, have been well documented
my (PD) and appear to be the result of partial gastrectomy and pancreatic exocrine insufficiency [1] . Atrophy of the re mnant distal pancreas is also frequen t and might be caused by ph ys iological ch ange in normal digestive func-
and relate to th e preservation of gastric capacity [3,4]. Several investigators h ave sh own that PPPD preserves the secretion of gastrointestinal (GI) hormones, such as gastrin, secretin and cholecystokinin (CCK) from the dis-
tion, or poor pancreatic drainage through the site of the pancreatoenteric anastomosis.
tal stomach and proximal duodenum; after PD their secretion almost completely stops [5- 8]. These differences could
Pylorus-preserving pan creatoduoden ectomy (PPPD) was first used in cases of chronic pancreatitis [2]. More
be associated with better nutrition after PPPD. It h as been suggested that some GI hormones have a trophic effect on the mucosa of the GI tract and pancreas
recently, PPPD h as tended to replace the con vention al Correspondence to: SW Kim, Deportment of Surgery, Seoul Notional University College of Medicine, 28 Yongon-Dong, Chongro-Ku, Seoul I 10-744, Korea
© I999 Isis Medical Media Ltd.
65
SW Kim et of. [9] and, in animal models, this physiological effect is well established [9-14]. Hypertrophic gastric mucosa in gastrinoma patients [15] provides good evidence of the trophic effect of GI hormones in humans. However, since it is almost impossible to establish a human model, the existence of this trophic effect has not been established . We postulated that the ablation of GI hormones, including gastrin and CCK, during PD is one of the important causes of postoperative pancreatic atrophy. Unlike PD, PPPD to some extent preserves these hormones, and so postoperative atrophy could be less marked. Better nutrition after PPPD would thus be due to the maintenance of pancreatic volume, as well as preservation of the pylorus.
cm of distal jejunum were anastomosed to the stomach or duodenum [16].
Pancreas volumetry Thin-section spiral CT was performed on the remnant distal pancreas; reconstructed images were used for 3D display and volumetric measurement. After localisation of the pancreas by conventional CT, spiral CT was performed on the region of the pancreas 30-65 s after the injection of i.v. contrast material at a speed of 3 cm 3 S- 1 and images were reconstructed at 5 mm intervals. The reconstructed CT image data were sent to a 3D graphic work-station (Allegro, ISG Technologies, Missisauga, C anada). On the work-station, a threshold was
To determine whether there is evidence for this hypothesis, postoperative pancreatic volume and stimulat-
set to include the area of the pancreas. Soft tissue areas other than the pancreas that were included by the thresh-
ed GI hormone release were measured in patients who underwent either PD or PPPD for periampullary cancer.
old were manually removed from each image by the abdominal radiologist. After the editing, a 3D object was generated and the volume was calculated by the work-station. To minimise the possible variation caused by the threshold setting, the threshold value was set to be constant in all
Methods Patients Patients who underwent PPPD (n
=
16) or PD (n
=
12) for
periampullary cancer performed by the same surgeon and who survived 2 1 year without evidence of recurrence were enrolled in this study. PPPD was the operation of choice; PD was performed if the lesion was of pancreatic origin, if the patient had gastric lesions, such as peptic ulcer or gastritis (either at the time of diagnosis or based on the medical history), or if the blood supply of the duodenum looked insufficient during PPPD. Patients with preoperative pancreatic atrophy were excluded.
Surgical techniques During PD or PPPD the pancreas was cut along the midline of the portal and superior mesenteric vein. Either pancreatogastrostomy (PG, n = 15) or pancreatojejunostomy (PJ, n = 13) was used for reconstruction. The pancreas was anastomosed and invaginated into the antrum or the low gastric body (posterior wall) for PG and into the antimesenteric side of the jejunum for PJ. Anastomosis was performed using a two-layer interrupted suture with 4-0 black silk; a polyethylene tube about 3 cm long was inserted into the pancreatic duct as a stent. The extent of gastric resection during PD was hemigastrectomy, which would not affect the choice of PG. The proximal jejunal end was closed and an end-to-side hepaticojejunostomy was performed; 40-50
66
patients, between 14- 175 Hounsfield units. If preoperative spiral CT was available, this too was measured, as was the pancreatic volume distal to the mid-line (portal and superior mesenteric vein).
Gastrin and CCK assay After patients fasten for at least 8 h, a blood sample was taken before and at 15, 30, 45 and 60 min after ingestion of an amino acid-enriched fluid diet (EnsureR 1 can; 200 ml, 400 k cal). Samples were collected in Trasylol and EDTA-treated tubes. Plasma was taken after centrifuge and was stored in the freezer until hormonal assay. Radioimmunoassays, using antiCCK antibody (OAL-656, Ostuka Assay Laboratory, Japan) and a gastrin kit (DPC, USA) were performed [17].
Evaluation of nutritional status Body weight change and current body mass index (BMI) were compared between PD and PPPD groups. The extent of recovery to the body weight on admission was stratified as 2 100%,95- 100% and < 95%. Body weight heighC 1 (kg m- 2 ) was used as BMI.
Statistical evaluation The results are expressed as mean ± SE and an independent t test was used to evaluate statistical significance. p < 0.05 was considered significant.
PPPD prevents pancreatic atrophy
Results The patients' clinical and nutritional profile, according to the type of operation, is shown in Tables 1 and 2. The nutritional status with respect to weight recovery and BMI was better in the PPPD group. However, these differences were only of the p = 0.07 and 0.06 level. The mean volume of the remnant distal pancreas after PPPD (21 S4S±2117 mm3 ) was greater than that after standard PD (13 Sl1 ±1219 mm 3 ) (p < 0.05). Figure 1 sh ows the mean volume of the remnant distal pancreas per body weight in the two groups of patients. Pancreatic volume per body weight after PPPD (388.0±39.9 mm3 kg- I) was greater than that after PD (263.3±27.3 mm3 kg- I) (p < 0.05) . Preoperative volumetry of the pancreas distal to the mid-line of the superior mesenteric-portal vein could be performed in 12 patients. There was no difference in volume per body weight between the PPPD group (n = 7; 771.4 ±122 .S mm 3 kg- I) and the PD group (n = 5; 727.S ±20S.6 mm3 kg- I). The extent of volume decrease, however, was much less in the PPPD group (42.9%) than in the PD group (69.4%) (p = 0.024) (Figure 2).
Basal and stimulated gastrin release in the PPPD group and standard PD group is shown in Figure 3. In the PPPD group, plasma gastrin increased from S0.3±6.7 pg ml- I to 68.6 ±6.0 pg ml- 1 at 15 min after stimulation, wh ereas in the standard PD group there was no stimulated gastrin release and throughout the time course gastrin levels were significantly lower than in the PPPD group. There was a significant correlation between the gastrin level at 15 min after stimulation and the volume of remnant pancreas (Pearson correlation coefficient, p < 0.05) (Figure 4). Basal CCK level was 3.6±1.3 pg mL- I in the standard PD group and 4.8±1.1 pg ml- 1 in the PPD group. 30 min after stimulation, peak CCK response was 7.7 ±2.1 pg mL- 1
Cfl
1 3 800 mm kg-
CIl Q)
U c 700 CIl Q.
600
C
CIl
c
500
~
400
E
• •
v----1~
·
Q)
-£ 300 '0 200 Q)
E 100 ::J
~
Table I. Patient details
= 12)
PO (n
PPPO (n
= 16)
•• • • •
•
0
PO (n=12)
PPPO (n = 16)
Figure I. Volume of the pancreas after each type of pancreatoduodeaomy. =0.017.
p
Mean age (years) Male : Female Disease entity Ampulla of Vater cancer Common bile duct cancer Pancreatic cancer Duodenal cancer PJfPG anastomosis Postoperative month (mean)
59.7 2:I
61.7 I: I
2 7 2
II
70
5/7 23.2 (12-46)
5 0 0 8/8 32.9 (13-65)
60 Q)
~
.>~ _ 0
'§ --; 50 Q) CIl Q.Q)
g .....
g40
CIl -2-Q. Q)
Q)
.,E; ~30 CIlIDO
.8 E20
Table 2. Body weight recovery after operation Body weight recover
PO
(n %ratio to body weight on admission ~ 100% 95-100% < 95% BMI (kg m-2)
= 12)
PPPO (n = 16)
Cfl ::J 0_
P
Cl... 0
> 10
o
PO
Preoperative volume : 2 (16.7) 7 (43.8) I (8.3) 3 (18.8) 9 (75.0) 6 (37.5) 20.5 ± 0.9 22.7 ± 0.7
727.5+205.6 mm3 kg-1 771.4+122 .5 mm3 kg-1 n=7 0.074 0.064
n=7
Figure 2. The volume change of distal pancreas after each type of pancreatoduodeaomy. Preoperative volume includes the pancreas distal to the midline of the portal and superior mesenteric vein. p = 0.024.
67
SW Kim et 01.
~
E
80
18
70
16
60
14
OJ
'I
50 E:: c
enC\l
40
PPPD (n = 16)
OJ
C\l
30
C\l
20
E (fJ 0:::
12
--l
.;::
i-
~.
0..
~ 0 0
PD(n=12)
8
C\l
6
C\l
4
E (fJ
.....
10
E 10 OJ
0::: 2
0
15
0
30
45
60
0
Min
0
15
30
45
60
Min Figure 3. Secretion
of gastrin after each type of pancreatoduodectomy.
p < 0.005.
Figure 5. Patterns of CCK secretion after each type of pancreatoduodectomy. The difference in stimulated CCK release was not statistically significant.
c 120 (pg mL- 1) 0
~
•
::l E 100
.~
CD
80
c
60
• •
~
'E
who have undergone PPPD than in th ose who have undergone conventional Po. Since postoperative atrophy of the remnant pancreas is not exceptional, we interpre t this observation as showing that there is less atrophy after PPPD
• •
than after PD. Preoperative volumetry was performed in only 12 of 26 patients. The CT scans of many patients
L!)
Qi
40
•
>
.S!? c
20
"5 (fJ C\l
C)
00
100
...'
• • 200
300
Pancreatic
• 400
• 500
600
700
800
volume(mm 3 kg- 1)
Figure 4. Correlation between gastrin level 15 min after stimulation and volume of the remnant pancreas. There was significant correlation between these two factors (Pearson correlation coefficient =0.4815, P = 0.013).
in the standard PO group and 13.3±3.9 pg mL- i in the PPPD group. The stimulatory response of CCK was higher in the PPPD group than in the PO group, though the difference was not statistically significant (Figure 5). There was no correlation between CCK and pancreatic volume. Other factors that might have affected postoperative pancreative volume, namely type of reconstruction (PG versus PJ), age, postoperative interval and nutritional status (BM!) were investigated (Table 3). None correlated with pancreatic volume.
referred to us were no longer available, so their preoperative pancreatic volume could not be accurately checked. In this small group of 12 patients, however, there was a significant difference in postoperative volume of the pancreas, even though their preoperative volumes were similar. Better nutritional status after PPPD, as indicated by weight recovery and other parameters, h as been reported [1,2]. In this series, however, nutritional status between PPPD and PO were not significantly different although a trend in favour of PPPD was apparent. Nor was pancreatic volume per weight or body surface area significantly different; thus, nutritional factors do not appear to h ave any effect. Mainz and colleagues [10] were the fi rst to show that G I hormones , including CCK, increased pancreatic DN A, thus stimulating the growth of this organ. N umerous studies [9-12] have documented that gastrin and CCK or its an alogue stimulate the growth of the exocrine pancreas in
Discussion
different animal species, but this effect has not yet been demonstrated in man. Since long-term manipulation with exogenous or endogenous GI hormones is almost impossi-
This study has demonstrated, for the first time, rhat rhe \-01ume of the distal remnant pancreas is greater in patients
ble in men , suitable experiments would be very difficult to conduct.
68
PPPD prevents pancreatic atrophy
Table 3. Pan creative volume and clinical factors
Type of operation
PPPD PD
Pancreatic anastomosis
PJ PG :S;60
Age (years) Sex Postoperative adjuvant treatment Postoperative interval (months)
>60 Male Female Yes No :s;24 >24
n patients
Pancreas volume (mm 3 kg~') Mean±SE
16 12 13 15 13 15 16 12 II 17 12 16
388.0±39.9 263.3±27.3 330.7±38.s 334.1 ±40.3 386.2±47.8 282.9±24.4 3Is.2±3s.6 3s4.3±44.4 283.3±26.3 366.6±41.6 331.7±SO.S 333.3±30.9
p 0.017 NS 0.070 NS 0.141 NS
NS = not significant. SE = standard error.
After PPPD, many patients maintain their normal GI hormone response to stimuli, whereas after Whipple operation the secretion of hormones, such as gastrin, CCK and secretin, is negligible [5-8]. GI hormones might thus exert a trophic effect on the human pancreas. This effect is a possible explanation for the increased pancreatic volume found after PPPD.
Late symptoms of gastrointestinal disturbance, including diarrhoea, are usually not serious. It is very difficult to differentiate the causes of diarrhoea, since dumping syndrome [4]' pancreatic exocrine insufficiency, or dissection of the nerve plexus around the celiac and superior mesenteric artery give rise to similar symptoms. In our series of
Such a model might demonstrate the trophic effect of GI hormones in man. The results of this study suggest only a possible relationship. We showed that basal and stimulated gastrin release was greater in the PPPD group of patients, as previous reports have demonstrated [6-8]. Interestingly, there was a sig-
patients, only a few complained of intermittent diarrhoea, which was not clinically important and did not differ
nificant correlation between stimulated gastrin release and the volume of remnant pancreas, which suggests that the
well known, but whether pancreatic regeneration occurs after partial resection is uncertain. It has been demonstrated in an animal model that partial pancreatectomy stimulates remnant pancreatic growth, and that increased
extent of hormonal preservation may be related to the preservation of function and the volume of the target organ. Mean secretory CCK response was higher after PPPD, though CCK release was not seen in all cases. Whether
according to the type of resection or whether reconstruction was by PJ or PG. Regeneration of the liver after partial liver resection is
endogenous CCK enhances the regenerative effect [13,14]. Whether this same effect occurs in man remains uncertain.
CCK release is preserved depends, in part, on the length and function of the preserved first portion of the duodenum, which can vary from patient to patient. CCK can also
It may be assumed that distal pancreatectomy stimulates growth of the pancreatic head, whereas resection of the head, together with duodenum, might not stimulate growth
be released from the proximal jejunum and this fact can explain its release after PD. Pancreatic atrophy is a chronic postoperative change
of the distal pancreas because of the ablation of CCK release.
which can differ according to the interval between operation and the measurement of pancreatic volume. After
Underlying pancreatic pathology may affect the rate of proliferation of the remnant pancreas, though it is not clear
each operation type, however, the interval was similar and there was no correlation between length of interval and
whether chronic inflammation in the pancreas affects the extent of atrophy. In our series of patients, moderate-tosevere chronic inflammation in the margin of the resected
pancreatic volume.
pancreas was not detected.
69
SW Kim et 01. After pylorus preservation, originally performed to prevent the postgastrectomy syndrome, nutrition is more satisfactory [18] and this may also explain why there is less atrophy after PPPD. In this study, however, there was no differ-
6
7
ence in nutritional status and the absolute volume of the remnant pancreas did not correlate with nutritional status, BMI or body weight. In conclusion, the greater volume of the remnant pancreas after PPPD rather than PD may be due to the preservation of GI hormones, such as gastrin and CCK, that exert a trophic effect on the pancreas. As suggested in our previous report [16] the stimulatory effect of GI hormones on growth may affect anastomotic wound healing. Pyloric preservation results in the continuing secretion of GI hormones, thereby reducing anastomotic leakage after pancreatoduodenectomy.
Acknowledgement
8
Surg 1991;214:56-60. 9
10 11
12
Supported by grant number 95-010 from Seoul National University Hospital.
References
2
3
4
5
70
Traverso LW, Longmire WP. Preservation of the pylorus in pancreaticoduodenectomy. A follow-up evaluation. Ann Surg 1980;192:306-10. Traverso LW, Longmire WP. Preservation of the pylorus during pancreaticoduodenectomy. Surg Gynecol Obstet 1978; 146:959-62. Morel P, Mathey P, Corboud H et al. Pylorus-preserving duodenopancreatectomy: long-term complications and comparison with the Whipple procedure. World] Surg 1990;14:642-7. Braasch JW, Rossi RL, Watkins E Jr, Danniel DJ. Pyloric- and gastric-preserving pancreatic resection: experience with 87 patients. Ann Surg 1986;204:411-18. Pearlman NW, Stiegman GV, Ahnen DJ et al. Acid and gastrin levels following pyloric-preserving pancreaticoduodenectomy. Arch Surg 1986;121:661-4.
Saro T, Imamura M, Matsuno S et al. Gastric acid secretion and gut hormone release in patients undergoing pancreaticoduodenectomy. Surgery 1986;99:728-34. Takada T, Yasuda H, Shikaka J et al. Postprandial plasma gastrin and secretin concentrations after a pancreatoduodenectomy: a comparison between a pylorus-preserving pancreatoduodenecromy and the Whipple procedure. Ann Surg 1989;210:47-51. Tangoku A, Nishikawa M, Adachi A, Suzuki T. Plasma gastrin and cholecystokinin response after pylorus preserving pancreatoduodenectomy with Billroth-I type of reconstruction. Ann
13
Johnson LR. Regulation of gastrointestinal growth. In: Johnson DR editor. Physiology of the gastrointestinal tract. 2nd ed. New York: Raven Press, 1987:301-3. Maim DL, Black 0, Webster PD. Hormonal control of pancreatic growth.] Clin Invest 1973;52:2300-4. Dembinski AB, Johnson LR. Stimulation of pancreatic growth by secretin, caerulein and pentagastrin. Endocrinology 1980;105:323-8. Morriset J, Chamberland S, Gilbert L et al. Study of pancreatic DNA synthesis in vivo and in vitro following caerulein treatment in vivo. Biomed Res 1982;3:151-8. Johnson LR, Dudrick SJ, Guthrie PD. Stimulation of pancreatic growth by intraduodenal amino acids and HC!. Am ]
PhysioI1980;239:G400-5. 14 Lehv M, Fitzgerald PJ . Pancreatic acinar cell regeneration. IV. Regeneration after surgical resection. Am ] Pathol 1986;53:513-35. 15 Neuburger P, Lewin M, de Recherche C et al. Parietal and chief cell populations in four cases of Zollinger-Ellison syndrome. Gastroenterology 1972;63:937-42 16 Kim SW, Youk EG, Park YH. Comparison of pancreatogastrostomy and pancreatojejunostomy after pancreatoduodenectomy performed by one surgeon. World] Surg 1997;21:640-2. 17 Hashimura E, Shimizu F, Nishino T et al. Production of rabbit antibody specific for aminoterminal residues of cholecystokinin octapeptide (CCK-8) by selective suppression of crossreactive antibody response. ] Immunol Method 1982;5:375-87. 18 Linehan IP, Russell RCG, Hobsley M. The dumping syndrome after pancreaticojejunostomy. Surg Gynecol Obstet 1988;167: 114-18.