- Email: [email protected]
Pancreaticojejunal Anastomosis Is Preferable to Pancreaticogastrostomy after Pancreaticoduodenectomy for Longterm Outcomes of Pancreatic Exocrine Function
Pancreaticojejunal Anastomosis Is Preferable to Pancreaticogastrostomy after Pancreaticoduodenectomy for Longterm Outcomes of Pancreatic Exocrine Function Alexandre Rault, MD, Antonio SaCunha, MD, Daniel Klopfenstein, MD, Dominique Larroudé, MD, Frédéric N Dobo Epoy, MD, Denis Collet, MD, Bernard Masson, MD The aim of this study was to evaluate pancreatic exocrine and endocrine function after pancreaticoduodenectomy. STUDY DESIGN: Pancreatic exocrine function was evaluated by a questionnaire and medical examination of stools after discontinuing pancreatic enzyme supplements for at least 10 days. Severe steatorrhea was defined as frequent, nauseating, yellow, and pasty stools, fecal output ⬎ 200 g/d for more than 3 days. Endocrine function was evaluated by blood glucose level. Association between severe steatorrhea and age, indication, histologic obstructive pancreatitis, pancreaticojejunal anastomosis (PJA), pancreaticogastric anastomosis (PGA), and morbidity was studied. RESULTS: Fifty-two patients underwent pancreaticoduodenectomy, complication rate was 33%. PJA was performed in 41 patients (79%) and PGA in 11 patients (21%). At a median followup of 75 months (24 to 156 months), 65% of the patients received pancreatic enzyme supplements. Severe steatorrhea was observed in 22 patients (42%). Incidence of postoperative diabetes was 14.6%. Patient age (more than 60 years), postoperative complication, and obstructive pancreatitis were not associated with postoperative severe steatorrhea. In cases of nonhistologic obstructive pancreatitis, PGA was more frequently associated with severe steatorrhea than PJA (70% versus 21.7%, p ⬍ 0.025). No factor significantly influenced incidence of postoperative diabetes. CONCLUSIONS: After pancreaticoduodenectomy, 42% of patients presented with severe steatorrhea. PJA allows better pancreatic exocrine function preservation than PGA and should be recommended. (J Am Coll Surg 2005;201:239–244. © 2005 by the American College of Surgeons) BACKGROUND:
pancreatic leakage, which is the most serious complication observed after pancreaticoduodenectomy. A randomized controlled study comparing PJA and PGA did not show any notable difference in terms of complications2 and few studies reported any difference in longterm outcomes of pancreatic function. In this study, my colleagues and I compared longterm functional outcomes of pancreaticoduodenectomy, depending on the type of pancreaticoenterostomy: pancreaticojejunostomy or pancreaticogastrostomy.
The Whipple procedure, which has been the standard treatment for periampullary carcinomas1 and pancreatic cancer, is still associated with high operative morbidity and mortality. Numerous modifications have been developed to reduce operative risk. For example, after pancreaticoduodenectomy, gastrointestinal pancreatic drainage can be restored using either a pancreaticojejunal anastomosis (PJA) or a pancreaticogastric anastomosis (PGA). Pancreaticogastrostomy was introduced to reduce
METHODS
Competing Interests Declared: None.
Patients
Received January 30, 2005; Accepted March 18, 2005. From CHU Bordeaux, Maison du Haut-Lévéque, Service de Chirurgie Digestive, Av Magellan 33604 Pessac Cedex, France. Correspondence address: Alexandre Rault, CHU Bordeaux, Maison du Haut-Lévéque, Service de Chirurgie Digestive, Av Magellan, F 33604 Pessac Cedex, France.
© 2005 by the American College of Surgeons Published by Elsevier Inc.
All patients who underwent pancreaticoduodenectomy at Haut Lévéque Hospital, CHU Bordeaux, France between November 1989 and December 2001 were eligible for this study. During this period 272 procedures
239
ISSN 1072-7515/05/$30.00 doi:10.1016/j.jamcollsurg.2005.03.026
240
Rault et al
Pancreaticojejunal Anastomosis
were performed for various indications. We excluded patients with chronic calcified pancreatitis, patients addicted to alcohol, and patients who had previously undergone a procedure that might alter bowel absorption (eg, colectomy, small bowel resection, and total or partial gastrectomy) because of difficulties interpreting pancreatic endocrine and exocrine function. We also excluded patients whose remnant pancreas was treated by chemical occlusion (n ⫽ 3). Statistical analysis was performed in April 2004, allowing followup of at least 24 months. Only 52 patients were finally eligible for this study: 61 patients were excluded and 159 patients were lost to followup or died. Remnant pancreas was macroscopically studied and observed by the surgeon and presence of obstructive pancreatitis was noted. A pathologic examination of the pancreatic section looked for histologic obstructive pancreatitis for all patients and for malignancy in cases of pancreatic adenocarcinoma. The operative procedure included antrectomy. The distal pancreas was anastomosed to the antral posterior wall of the stomach for the PGA and to the antimesenteric side of the first jejunal loop for the PJA using ductto-mucosa methods. The biliary continuity was realized by an end-to-side hepaticojejunal anastomosis on the first jejunal loop, 50 to 60 cm proximal to the gastrojejunal anastomosis. The procedures performed did not include pylorus-preserving pancreaticoduodenectomy. Factors influencing the presence of postoperative severe steatorrhea were studied: age, gender, surgical morbidity, indication for resection, type of anastomosis (pancreaticojejunal, pancreaticogastric), histologic obstructive pancreatitis, and preoperative radiotherapy. Evaluation of pancreatic exocrine function
For this study, we used a clinical evaluation of the pancreatic exocrine function. A questionnaire about the appearance of stools was sent to all the patients in the study. Questions were asked about number of stools per day, smell and color of the stools, and amount of fecal output. Stools were also examined by a single trained doctor who collected all these criteria. We also studied variation in body weight before and after the procedure. Patients who were taking pancreatic enzyme supplements were asked to discontinue this treatment at least 10 days before the clinical evaluation of their stools. A clinical steatorrhea was defined by more than three stools per day, fecal output of ⬎ 200 g/d for at
J Am Coll Surg
least 3 consecutive days, nauseating smell, pale or yellow stools, and the appearance of stools as pasty or greasy. A severe steatorrhea was defined by presence of at least three of these criteria. Evaluation of pancreatic endocrine function
To assess endocrine function, fasting blood glucose level (normal range ⬍ 110 mg/dL) was measured without administration of an oral hypoglycemic agent or insulin. A diagnosis of diabetes mellitus was made based on criteria set by the 1985 World Health Organization study group on diabetes mellitus.3 Statistical analysis
An analysis between matched variables was performed using paired Student’s t-test and chi-square test for qualitative parameters. A value of p ⬍ 0.05 was considered significant. RESULTS With regard to general parameters, including age, gender, and interval time of operation and study, the groups were comparable. Indications for resection were also comparable in each group. Patient characteristics: indication for pancreaticoduodenectomy
Median age was 61 years (range 27 to 79 years), there were 27 men and 25 women. Median interval between operation and evaluation was 75 months (range 24 to 156 months) without any notable difference between groups of patients. Indications for resection were malignant diseases (n ⫽ 32), including 9 pancreatic adenocarcinomas, 11 ampullary adenocarcinomas, 5 endocrine tumors, 4 duodenal tumors (1 gastrointestinal stromal tumor and 3 adenocarcinomas), and 3 cholangiocarcinomas. Other indications for pancreatectomies were 4 intrapapillary mucinous tumors (all localized), 2 inflammatory stenosis of the common bile duct, 3 ampullary tumors, 6 cystic tumors, and 5 focused chronic pancreatitis (Table 1). Histology of pancreatic resection margins was without malignancy in all patients. Four patients received preoperative radiotherapy (for pancreatic adenocarcinomas). Results of the operation
Complications occurred in 17 patients (33%): 12 pancreatic leakages (23%) (1 reoperation), 1 biliary leakage
Vol. 201, No. 2, August 2005
Rault et al
Table 1. Patient Characteristics Characteristics
Median age (y), range Gender ratio (male to female) Median followup (mo), range Indication for resection, n (%) Malignant Benign
Pancreaticojejunal Anastomosis
241
Table 2. Results in Nonobstructive Pancreatitis Patients
61, 27–79 2.03 75, 24–156 32 (61.5) 20 (38.5)
(2%) (medically treated), and 4 delayed gastric emptying (8%). Presence of a histologic obstructive pancreatitis did not considerably decrease risk of pancreatic fistula. No notable difference in pancreatic leakage was observed between PJA and PGA. Pancreatic exocrine function
With median followup of 75 months after operation, 34 of the 52 patients (65%) received enteric-coated pancreatic supplements. These supplements were generally prescribed by a general practitioner because of diarrhea. Pancreatic enzyme supplements were discontinued before stool collection. Twenty-two patients (42%) presented with severe steatorrhea. Nearly 50% of the patients presented with weight loss ⬎ 10% of their preoperative weight. We observed a substantial difference in the pasty stool appearance between PJA and PGA in patients without histologic obstructive pancreatitis (21.7% versus 70%) (p ⬍ 0.025). A substantial difference was also observed in severe steatorrhea between these two types of anastomosis in patients without histologic obstructive pancreatitis: the severe steatorrhea is considerably more frequent in the PGA group than in PJA group in patients without histologic obstructive pancreatitis (p ⬍ 0.025) (Table 2 to 3). No marked difference was observed between patients in presence or absence of histologic obstructive pancreatitis concerning postoperative exocrine pancreatic insufficiency. Histologic obstructive pancreatitis did not seem to be a risk factor for exocrine dysfunction after pancreaticoduodenectomy. All patients who received preoperative radiotherapy presented histologic obstructive pancreatitis and pancreatic exocrine insufficiency. Preoperative radiotherapy, age, the proportion of postoperative complications, or the indication for resection did not substantially increase the risk of postoperative pancreatic insufficiency. Presence of histologic obstructive pancreatitis did not bring about postoperative severe steatorrhea, whichever type of anastomosis was used (PJA or PGA) (Table 4).
Weight loss PES Pasty stools Severe steatorrhea
Nonhistologic obstructive pancreatitis (n ⴝ 33) PJA PGA (n ⴝ 23) (n ⴝ 10) n % n %
p Value
12 12 5 5
NS NS ⬍ 0.025 ⬍ 0.025
52.2 52.2 21.7 21.7
5 8 7 7
50 80 70 70
NS, not significant; PES, pancreatic enzyme supplements; PGA, pancreaticogastric anastomosis; PJA, pancreaticojejunal anastomosis.
Pancreatic endocrine function
Of patients with histologic obstructive pancreatitis, 26.7% presented preoperatively with glucose intolerance, although this rate was only 6.1% in the absence of histologic obstructive pancreatitis (p ⬎ 0.05). In the total cohort of patients, the incidence of postoperative glucose intolerance was 14.6% (16.1% in patients with histologic obstructive pancreatitis and 9.1% in the others). There was no statistically significant difference regarding the incidence of abnormal glucose tolerance between patients with or without histologic obstructive pancreatitis, or between type of pancreatic anastomosis (with or without histologic obstructive pancreatitis). Glucose tolerance improved in one patient in the malignant group (pancreatic adenocarcinoma) after operation. DISCUSSION There are few studies comparing the functional results of PGA and PJA in patients who have undergone pancreaticoduodenectomy. Theoretically, PGA could cause more functional derangement because reflux of gastric juices causes inactivation of the pancreatic enzymes and early insufficiency of the remnant pancreas. Shinchi and Table 3. Results in Obstructive Pancreatitis Histologic obstructive pancreatitis (n ⴝ 19) PGA PJA (n ⴝ 18) (n ⴝ 1) n % n
Weight loss PES Pasty stools Severe steatorrhea
8 12 9 9
57.1 66.6 50 50
— 1 1 1
p Value
NS NS NS NS
NS, not significant; PES, pancreatic enzyme supplements; PGA, pancreaticogastric anastomosis; PJA, pancreaticojejunal anastomosis.
242
Rault et al
Pancreaticojejunal Anastomosis
Table 4. Criteria Influencing Severe Steatorrhea
Criteria
Age (y) ⬍ 60 ⬎ 60 Gender Male Female Indication for operation Malignant Benign Morbidity Yes No HOP Yes No Anastomosis PJA PGA Preoperative radiotherapy PJA (n ⫽ 4) Yes No Non-HOP (n ⫽ 33) PJA PGA HOP (n ⫽ 15) PJA PGA PJA (n ⫽ 41) HOP Non-HOP PGA (n ⫽ 11) HOP Non-HOP
Severe Nonsevere p Value steatorrhea steatorrhea (severe (n ⴝ 22) (n ⴝ 30) steatorrhea)
8 14
7 23
⬎ 0.05
13 9
12 18
⬎ 0.05
14 8
18 12
⬎ 0.05
5 17
12 18
⬎ 0.05
10 12
9 21
⬎ 0.05
14 8
27 3
⬎ 0.05
4 18
0 30
5 7
18 3
5 1
9 0
9 5
9 18
1 7
0 3
—
⬍ 0.025
—
⬎ 0.05
—
HOP, histologic obstructive pancreatitis; PGA, pancreaticogastric anastomosis; PJA, pancreaticojejunal anastomosis.
colleagues4 recommended pancreaticogastrostomy as a physiologic reconstruction procedure after pancreaticoduodenectomy because of its maintenance of normal gastric pH and its preservation of the neurohumoral relation between the stomach, duodenum, and pancreas. This was deduced after pancreaticogastrostomy by measuring gastric acid secretion, serum gastric levels, and 24-hour dual gastroduodenal pH. Takada and colleagues5 investigated patients who underwent pancreaticoduodenectomy with pancreaticogastrostomy. Although the stool p-type amylase, lipase, and chymotrypsin activity in the
J Am Coll Surg
PGA group amounted to 21%, 27%, and 31% of respective values seen at normal levels, they did not differ substantially from the levels seen in the PJA group. In our study, with a median followup of 75 months, 65% of patients received enteric-coated pancreatic supplements. This was similar to previously reported studies. There was a marked difference (p ⬍ 0.025) in patients with severe steatorrhea between PJA and PGA in the group without histologic obstructive pancreatitis. Because there was no marked difference between PJA and PGA in the group with histologic obstructive pancreatitis, histologic obstructive pancreatitis is not a factor influencing postoperative functional results after pancreaticoduodenectomy. After the tumoral obstacle has been resected, the pancreatic parenchyma recovers his preoperative function and the postoperative exocrine function is not markedly different from patients without histologic obstructive pancreatitis. Preoperative radiotherapy does seem to be a risk factor for exocrine dysfunction after pancreaticoduodenectomy. There were few studies with followup of 75 months, of exocrine function after pancreaticoduodenectomy, and few studies showed any difference in the exocrine function between PJA and PGA.6 Jang and colleagues7 studied 34 patients after pancreaticoduodenectomy with PGA (14 patients) and PJA (20 patients). There were 4 mild and 15 severe cases of pancreatic exocrine insufficiency among those who underwent PJA, but all patients in the PGA group showed severe pancreatic insufficiency (p ⫽ 0.045). The authors concluded that PGA is a safer anastomotic procedure in terms of leakage after pancreaticoduodenectomy, but it was associated with more functional deterioration. Mean followup for this study was only 24 months. For this study, we have chosen a clinical evaluation for steatorrhea. According to a report by Lankisch and colleagues,8 the correct visual diagnosis rate for feces in normal subjects with fecal fat excretion not exceeding 7 g/dL was 94%. In severe steatorrhea, where fecal fat excretion exceeded 15 g/d, the correct diagnosis rate was 82%. In mild steatorrhea, where fecal fat excretion was between 7 and 15 g/d, the correct diagnosis rate went as low as 54%. In this study, the correct diagnosis rate for steatorrhea seemed to improve when appearance and odor of feces, in addition to fecal mass, are considered. These results have been confirmed by Nakamura and colleagues,9 whose study showed that diagnosis of steatorrhea based on visual observation of fecal mass and
Vol. 201, No. 2, August 2005
odor gave favorable results: sensitivity was 89.3% and specificity was 91.1%. Fecal-1 elastase allows indirect diagnosis of pancreatic exocrine dysfunction with a sensitivity and specificity of 93% and 93%, respectively.10-12 The author concluded that the correct observation of feces is the shortest path to the diagnosis of steatorrhea. We have also chosen this evaluation because frequency of stools and fecal mass are the direct symptoms felt by the patients, and this is a contribution to the patient’s good quality of life after pancreaticoduodenectomy. This result of exocrine pancreatic function is directly linked to quality of life after duodenopancreatectomy as demonstrated by Jang and colleagues7 in a study comparing PJA and PGA in 34 patients. This study showed that many patients reported gastrointestinal symptoms and pancreatic endocrine and exocrine deterioration, which lower quality of life and require frequent care (including therapy for diabetes). After pancreaticoduodenectomy, the wide extent of the resection itself brings about functional deterioration of the digestive system, including endocrine insufficiency. In our study, incidence of postoperative diabetes was only 14.6%, which can be compared with other studies in the literature.13,14 Despite the obvious reduction in insulin secretion, patients undergoing pancreatic resection for periampullary malignancy usually do not have overt diabetes and actually show a notable improvement of glucose metabolism.15-17 Based on studies on the relationship between diabetes and pancreatic cancer, it was hypothesized that profound peripheral insulin resistance caused by a tumor-associated diabetogenic factor may be responsible for the high frequency of diabetes in patients with pancreatic cancer.15-17 In our study, we observed one case of a total disappearance of diabetes after pancreatic cephalic resection for pancreatic adenocarcinoma. Operative morbidity and mortality after duodenopancreatectomy is decreasing with improvements in operative techniques and intensive care methods. Operative mortality is still about 5% to 8%.18,19 The leading cause of this high mortality is pancreatic fistula. Rate of pancreatic fistula after pancreaticoduodenectomy has been reported to be 5% to 19%.20,21 In an attempt to reduce the occurrence of pancreatic fistula, many modifications in the management of the remnant pancreatic stump have been developed. The standard Whipple procedure proposed a pancreaticojejunostomy and later pancreaticogastrostomy was reported. Many reports favoring pancreaticogastrostomy
Rault et al
Pancreaticojejunal Anastomosis
243
have appeared recently. Kim and colleagues22 studied 86 patients to compare pancreatic leakage after pancreaticoduodenectomy between PJA and PGA. They reported that pancreatic leakage occurred at rates of 15.8% and 2.1%, respectively. Fabre and colleagues23 studied 160 patients who underwent PGA. They reported that pancreatic leakage occurred in 2.5% of the patients and that postoperative mortality rate was 3%. They also reviewed the large amount of literature concerning PGA and reported that the rate of pancreatic leakage was 4% on average.22,23 In a prospective randomized trial of about 145 patients, the John Hopkins group2 reported no difference in leakage rate between PJA and PGA (11.1% versus 12.3%). In our study, no statistically significant difference was found between PJA and PGA. We also observed that no marked difference was noted between patients with or without histologic obstructive pancreatitis, regardless of type of anastomosis. In the John Hopkins hospital experience,2 in a univariate analysis, pancreatic soft tissue appeared to be a factor considerably increasing risk of pancreatic fistula. In summary, with a median followup of 75 months after operation, patients who underwent pancreaticoduodenectomy with PGA presented considerably more pancreatic exocrine insufficiency (p ⬍ 0.025) than patients with PJA if there were no histologic obstructive pancreatitis. Complication rates of these two procedures are not statistically different. We recommend, each time it is possible, to perform pancreatectomy with PJA anastomosis. REFERENCES 1. Whipple AO, Parsons WB, Mullins CR. Treatment of pancreatic carcinoma of the ampulla of Vater. Ann Surg 1935;102: 763–779. 2. Yeo CJ, Cameron JL, Maher MM, et al. A prospective randomized trial of pancreaticogastrostomy versus pancreaticojejunostomy after pancreaticoduodenectomy. Ann Surg 1995;222: 580–588. 3. Alberti KGMM, Hockaday TAR. Diabetes mellitus. In: Weatherall DJ, Ledingham JGG, Warell DA, eds. Oxford textbook of medicine. 2nd ed. New York: Oxford Medical Publications; 1987:51–101. 4. Shinchi H, Takao S, Fukura K, Aikou T. Gastric acid secretion and gastroduodenal pH after pancreaticogastrostomy in dogs. J Surg Res 1996;61:152–158. 5. Takada T, Yasuda H, Uchiyama K, et al. Pancreatic enzyme activity after a pylorus-preserving pancreaticoduodenectomy reconstructed with pancreaticogastrostomy. Pancreas 1995;11: 276–282.
244
Rault et al
Pancreaticojejunal Anastomosis
6. Andersen HB, Baden H, Brahe NE, Burcharth F. Pancreaticoduodenectomy for periampullary adenocarcinoma. J Am Coll Surg 1994;179:545–552. 7. Jang JY, Kim SW, Park SJ, Park YH. Comparison of the functional outcome after pylorus-preserving pancreatoduodenectomy: pancreatogastrostomy and pancreatojejunostomy. World J Surg 2002;26:366–371. Epub 2002 Jan 15. 8. Lankisch PG, Droge M, Hofses S, et al. Steatorrhoea: you cannot trust your eyes when it comes to diagnosis. Lancet 1996; 347(9015):1620–1621. 9. Nakamura T, Tando Y, Terada A, et al. Can pancreatic steatorrhea be diagnosed without chemical analysis? Int J Pancreatol 1997;22:121–125. 10. Loser C, Mollgaard A, Folsch UR. Faecal elastase 1: a novel, highly sensitive, and specific tubeless pancreatic function test. Gut 1996;39:580–586. 11. Stein J, Schoonbroodt D, Jung M, et al. Measurement of fecal elastase 1 by immunoreactivity: a new indirect test of the pancreatic function. Gastroenterol Clin Biol 1996;20:424–429. 12. Katschinski M, Schirra J, Bross A, et al. Duodenal secretion and fecal excretion of pancreatic elastase-1 in healthy humans and patients with chronic pancreatitis. Pancreas 1997;15:191–200. 13. Lemaire E, O’Toole D, Sauvanet A, et al. Functional and morphological changes in the pancreatic remnant following pancreaticoduodenectomy with pancreaticogastric anastomosis. Br J Surg 2000;87:434–438. 14. Van Berge Henegouwen MI, Moojen TM, van Gulik TM, et al. Postoperative weight gain after standard Whipple’s procedure versus pylorus-preserving pancreatoduodenectomy: the influence of tumour status. Br J Surg 1998;85:922–926.
J Am Coll Surg
15. Ahren B, Andren-Sandberg A. Capacity to secrete islet hormones after subtotal pancreatectomy for pancreatic cancer. Eur J Surg 1993;159:223–237. 16. Permert J, Ihse I, Jorfeldt L, et al. Improved glucose metabolism after subtotal pancreatectomy for pancreatic cancer. Br J Surg 1993;80:1047–1050. 17. Permert J, Adrian TE, Jacobsson P, et al. Is profound peripheral insulin resistance in patients with pancreatic cancer caused by a tumor-associated factor? Am J Surg 1993;165:61–66; discussion 66–67. 18. Baumel H, Huguier M, Manderscheid JC, et al. Results of resection for cancer of the exocrine pancreas: a study from the French Association of Surgery. Br J Surg 1994;81:102–107. 19. Trede M, Schwall G, Saeger HD. Survival after pancreatoduodenectomy. 118 consecutive resections without an operative mortality. Ann Surg 1990;211:447–458. 20. Wade TP, el-Ghazzawy AG, Virgo KS, Johnson FE. The Whipple resection for cancer in U.S. Department of Veterans Affairs Hospitals. Ann Surg 1995;221:241–248. 21. Cameron JL, Pitt HA, Yeo CJ, et al. One hundred and forty-five consecutive pancreaticoduodenectomies without mortality. Ann Surg 1993;217:430–435; discussion 435–438. 22. Kim SW, Youk EG, Park YH. Comparison of pancreatogastrostomy and pancreatojejunostomy after pancreatoduodenectomy performed by one surgeon. World J Surg 1997;21: 640–643. 23. Fabre JM, Arnaud JP, Navarro F, et al. Results of pancreatogastrostomy after pancreatoduodenectomy in 160 consecutive patients. Br J Surg 1998;85:751–754.
pancreatic leakage, which is the most serious complication observed after pancreaticoduodenectomy. A randomized controlled study comparing PJA and PGA did not show any notable difference in terms of complications2 and few studies reported any difference in longterm outcomes of pancreatic function. In this study, my colleagues and I compared longterm functional outcomes of pancreaticoduodenectomy, depending on the type of pancreaticoenterostomy: pancreaticojejunostomy or pancreaticogastrostomy.
The Whipple procedure, which has been the standard treatment for periampullary carcinomas1 and pancreatic cancer, is still associated with high operative morbidity and mortality. Numerous modifications have been developed to reduce operative risk. For example, after pancreaticoduodenectomy, gastrointestinal pancreatic drainage can be restored using either a pancreaticojejunal anastomosis (PJA) or a pancreaticogastric anastomosis (PGA). Pancreaticogastrostomy was introduced to reduce
METHODS
Competing Interests Declared: None.
Patients
Received January 30, 2005; Accepted March 18, 2005. From CHU Bordeaux, Maison du Haut-Lévéque, Service de Chirurgie Digestive, Av Magellan 33604 Pessac Cedex, France. Correspondence address: Alexandre Rault, CHU Bordeaux, Maison du Haut-Lévéque, Service de Chirurgie Digestive, Av Magellan, F 33604 Pessac Cedex, France.
© 2005 by the American College of Surgeons Published by Elsevier Inc.
All patients who underwent pancreaticoduodenectomy at Haut Lévéque Hospital, CHU Bordeaux, France between November 1989 and December 2001 were eligible for this study. During this period 272 procedures
239
ISSN 1072-7515/05/$30.00 doi:10.1016/j.jamcollsurg.2005.03.026
240
Rault et al
Pancreaticojejunal Anastomosis
were performed for various indications. We excluded patients with chronic calcified pancreatitis, patients addicted to alcohol, and patients who had previously undergone a procedure that might alter bowel absorption (eg, colectomy, small bowel resection, and total or partial gastrectomy) because of difficulties interpreting pancreatic endocrine and exocrine function. We also excluded patients whose remnant pancreas was treated by chemical occlusion (n ⫽ 3). Statistical analysis was performed in April 2004, allowing followup of at least 24 months. Only 52 patients were finally eligible for this study: 61 patients were excluded and 159 patients were lost to followup or died. Remnant pancreas was macroscopically studied and observed by the surgeon and presence of obstructive pancreatitis was noted. A pathologic examination of the pancreatic section looked for histologic obstructive pancreatitis for all patients and for malignancy in cases of pancreatic adenocarcinoma. The operative procedure included antrectomy. The distal pancreas was anastomosed to the antral posterior wall of the stomach for the PGA and to the antimesenteric side of the first jejunal loop for the PJA using ductto-mucosa methods. The biliary continuity was realized by an end-to-side hepaticojejunal anastomosis on the first jejunal loop, 50 to 60 cm proximal to the gastrojejunal anastomosis. The procedures performed did not include pylorus-preserving pancreaticoduodenectomy. Factors influencing the presence of postoperative severe steatorrhea were studied: age, gender, surgical morbidity, indication for resection, type of anastomosis (pancreaticojejunal, pancreaticogastric), histologic obstructive pancreatitis, and preoperative radiotherapy. Evaluation of pancreatic exocrine function
For this study, we used a clinical evaluation of the pancreatic exocrine function. A questionnaire about the appearance of stools was sent to all the patients in the study. Questions were asked about number of stools per day, smell and color of the stools, and amount of fecal output. Stools were also examined by a single trained doctor who collected all these criteria. We also studied variation in body weight before and after the procedure. Patients who were taking pancreatic enzyme supplements were asked to discontinue this treatment at least 10 days before the clinical evaluation of their stools. A clinical steatorrhea was defined by more than three stools per day, fecal output of ⬎ 200 g/d for at
J Am Coll Surg
least 3 consecutive days, nauseating smell, pale or yellow stools, and the appearance of stools as pasty or greasy. A severe steatorrhea was defined by presence of at least three of these criteria. Evaluation of pancreatic endocrine function
To assess endocrine function, fasting blood glucose level (normal range ⬍ 110 mg/dL) was measured without administration of an oral hypoglycemic agent or insulin. A diagnosis of diabetes mellitus was made based on criteria set by the 1985 World Health Organization study group on diabetes mellitus.3 Statistical analysis
An analysis between matched variables was performed using paired Student’s t-test and chi-square test for qualitative parameters. A value of p ⬍ 0.05 was considered significant. RESULTS With regard to general parameters, including age, gender, and interval time of operation and study, the groups were comparable. Indications for resection were also comparable in each group. Patient characteristics: indication for pancreaticoduodenectomy
Median age was 61 years (range 27 to 79 years), there were 27 men and 25 women. Median interval between operation and evaluation was 75 months (range 24 to 156 months) without any notable difference between groups of patients. Indications for resection were malignant diseases (n ⫽ 32), including 9 pancreatic adenocarcinomas, 11 ampullary adenocarcinomas, 5 endocrine tumors, 4 duodenal tumors (1 gastrointestinal stromal tumor and 3 adenocarcinomas), and 3 cholangiocarcinomas. Other indications for pancreatectomies were 4 intrapapillary mucinous tumors (all localized), 2 inflammatory stenosis of the common bile duct, 3 ampullary tumors, 6 cystic tumors, and 5 focused chronic pancreatitis (Table 1). Histology of pancreatic resection margins was without malignancy in all patients. Four patients received preoperative radiotherapy (for pancreatic adenocarcinomas). Results of the operation
Complications occurred in 17 patients (33%): 12 pancreatic leakages (23%) (1 reoperation), 1 biliary leakage
Vol. 201, No. 2, August 2005
Rault et al
Table 1. Patient Characteristics Characteristics
Median age (y), range Gender ratio (male to female) Median followup (mo), range Indication for resection, n (%) Malignant Benign
Pancreaticojejunal Anastomosis
241
Table 2. Results in Nonobstructive Pancreatitis Patients
61, 27–79 2.03 75, 24–156 32 (61.5) 20 (38.5)
(2%) (medically treated), and 4 delayed gastric emptying (8%). Presence of a histologic obstructive pancreatitis did not considerably decrease risk of pancreatic fistula. No notable difference in pancreatic leakage was observed between PJA and PGA. Pancreatic exocrine function
With median followup of 75 months after operation, 34 of the 52 patients (65%) received enteric-coated pancreatic supplements. These supplements were generally prescribed by a general practitioner because of diarrhea. Pancreatic enzyme supplements were discontinued before stool collection. Twenty-two patients (42%) presented with severe steatorrhea. Nearly 50% of the patients presented with weight loss ⬎ 10% of their preoperative weight. We observed a substantial difference in the pasty stool appearance between PJA and PGA in patients without histologic obstructive pancreatitis (21.7% versus 70%) (p ⬍ 0.025). A substantial difference was also observed in severe steatorrhea between these two types of anastomosis in patients without histologic obstructive pancreatitis: the severe steatorrhea is considerably more frequent in the PGA group than in PJA group in patients without histologic obstructive pancreatitis (p ⬍ 0.025) (Table 2 to 3). No marked difference was observed between patients in presence or absence of histologic obstructive pancreatitis concerning postoperative exocrine pancreatic insufficiency. Histologic obstructive pancreatitis did not seem to be a risk factor for exocrine dysfunction after pancreaticoduodenectomy. All patients who received preoperative radiotherapy presented histologic obstructive pancreatitis and pancreatic exocrine insufficiency. Preoperative radiotherapy, age, the proportion of postoperative complications, or the indication for resection did not substantially increase the risk of postoperative pancreatic insufficiency. Presence of histologic obstructive pancreatitis did not bring about postoperative severe steatorrhea, whichever type of anastomosis was used (PJA or PGA) (Table 4).
Weight loss PES Pasty stools Severe steatorrhea
Nonhistologic obstructive pancreatitis (n ⴝ 33) PJA PGA (n ⴝ 23) (n ⴝ 10) n % n %
p Value
12 12 5 5
NS NS ⬍ 0.025 ⬍ 0.025
52.2 52.2 21.7 21.7
5 8 7 7
50 80 70 70
NS, not significant; PES, pancreatic enzyme supplements; PGA, pancreaticogastric anastomosis; PJA, pancreaticojejunal anastomosis.
Pancreatic endocrine function
Of patients with histologic obstructive pancreatitis, 26.7% presented preoperatively with glucose intolerance, although this rate was only 6.1% in the absence of histologic obstructive pancreatitis (p ⬎ 0.05). In the total cohort of patients, the incidence of postoperative glucose intolerance was 14.6% (16.1% in patients with histologic obstructive pancreatitis and 9.1% in the others). There was no statistically significant difference regarding the incidence of abnormal glucose tolerance between patients with or without histologic obstructive pancreatitis, or between type of pancreatic anastomosis (with or without histologic obstructive pancreatitis). Glucose tolerance improved in one patient in the malignant group (pancreatic adenocarcinoma) after operation. DISCUSSION There are few studies comparing the functional results of PGA and PJA in patients who have undergone pancreaticoduodenectomy. Theoretically, PGA could cause more functional derangement because reflux of gastric juices causes inactivation of the pancreatic enzymes and early insufficiency of the remnant pancreas. Shinchi and Table 3. Results in Obstructive Pancreatitis Histologic obstructive pancreatitis (n ⴝ 19) PGA PJA (n ⴝ 18) (n ⴝ 1) n % n
Weight loss PES Pasty stools Severe steatorrhea
8 12 9 9
57.1 66.6 50 50
— 1 1 1
p Value
NS NS NS NS
NS, not significant; PES, pancreatic enzyme supplements; PGA, pancreaticogastric anastomosis; PJA, pancreaticojejunal anastomosis.
242
Rault et al
Pancreaticojejunal Anastomosis
Table 4. Criteria Influencing Severe Steatorrhea
Criteria
Age (y) ⬍ 60 ⬎ 60 Gender Male Female Indication for operation Malignant Benign Morbidity Yes No HOP Yes No Anastomosis PJA PGA Preoperative radiotherapy PJA (n ⫽ 4) Yes No Non-HOP (n ⫽ 33) PJA PGA HOP (n ⫽ 15) PJA PGA PJA (n ⫽ 41) HOP Non-HOP PGA (n ⫽ 11) HOP Non-HOP
Severe Nonsevere p Value steatorrhea steatorrhea (severe (n ⴝ 22) (n ⴝ 30) steatorrhea)
8 14
7 23
⬎ 0.05
13 9
12 18
⬎ 0.05
14 8
18 12
⬎ 0.05
5 17
12 18
⬎ 0.05
10 12
9 21
⬎ 0.05
14 8
27 3
⬎ 0.05
4 18
0 30
5 7
18 3
5 1
9 0
9 5
9 18
1 7
0 3
—
⬍ 0.025
—
⬎ 0.05
—
HOP, histologic obstructive pancreatitis; PGA, pancreaticogastric anastomosis; PJA, pancreaticojejunal anastomosis.
colleagues4 recommended pancreaticogastrostomy as a physiologic reconstruction procedure after pancreaticoduodenectomy because of its maintenance of normal gastric pH and its preservation of the neurohumoral relation between the stomach, duodenum, and pancreas. This was deduced after pancreaticogastrostomy by measuring gastric acid secretion, serum gastric levels, and 24-hour dual gastroduodenal pH. Takada and colleagues5 investigated patients who underwent pancreaticoduodenectomy with pancreaticogastrostomy. Although the stool p-type amylase, lipase, and chymotrypsin activity in the
J Am Coll Surg
PGA group amounted to 21%, 27%, and 31% of respective values seen at normal levels, they did not differ substantially from the levels seen in the PJA group. In our study, with a median followup of 75 months, 65% of patients received enteric-coated pancreatic supplements. This was similar to previously reported studies. There was a marked difference (p ⬍ 0.025) in patients with severe steatorrhea between PJA and PGA in the group without histologic obstructive pancreatitis. Because there was no marked difference between PJA and PGA in the group with histologic obstructive pancreatitis, histologic obstructive pancreatitis is not a factor influencing postoperative functional results after pancreaticoduodenectomy. After the tumoral obstacle has been resected, the pancreatic parenchyma recovers his preoperative function and the postoperative exocrine function is not markedly different from patients without histologic obstructive pancreatitis. Preoperative radiotherapy does seem to be a risk factor for exocrine dysfunction after pancreaticoduodenectomy. There were few studies with followup of 75 months, of exocrine function after pancreaticoduodenectomy, and few studies showed any difference in the exocrine function between PJA and PGA.6 Jang and colleagues7 studied 34 patients after pancreaticoduodenectomy with PGA (14 patients) and PJA (20 patients). There were 4 mild and 15 severe cases of pancreatic exocrine insufficiency among those who underwent PJA, but all patients in the PGA group showed severe pancreatic insufficiency (p ⫽ 0.045). The authors concluded that PGA is a safer anastomotic procedure in terms of leakage after pancreaticoduodenectomy, but it was associated with more functional deterioration. Mean followup for this study was only 24 months. For this study, we have chosen a clinical evaluation for steatorrhea. According to a report by Lankisch and colleagues,8 the correct visual diagnosis rate for feces in normal subjects with fecal fat excretion not exceeding 7 g/dL was 94%. In severe steatorrhea, where fecal fat excretion exceeded 15 g/d, the correct diagnosis rate was 82%. In mild steatorrhea, where fecal fat excretion was between 7 and 15 g/d, the correct diagnosis rate went as low as 54%. In this study, the correct diagnosis rate for steatorrhea seemed to improve when appearance and odor of feces, in addition to fecal mass, are considered. These results have been confirmed by Nakamura and colleagues,9 whose study showed that diagnosis of steatorrhea based on visual observation of fecal mass and
Vol. 201, No. 2, August 2005
odor gave favorable results: sensitivity was 89.3% and specificity was 91.1%. Fecal-1 elastase allows indirect diagnosis of pancreatic exocrine dysfunction with a sensitivity and specificity of 93% and 93%, respectively.10-12 The author concluded that the correct observation of feces is the shortest path to the diagnosis of steatorrhea. We have also chosen this evaluation because frequency of stools and fecal mass are the direct symptoms felt by the patients, and this is a contribution to the patient’s good quality of life after pancreaticoduodenectomy. This result of exocrine pancreatic function is directly linked to quality of life after duodenopancreatectomy as demonstrated by Jang and colleagues7 in a study comparing PJA and PGA in 34 patients. This study showed that many patients reported gastrointestinal symptoms and pancreatic endocrine and exocrine deterioration, which lower quality of life and require frequent care (including therapy for diabetes). After pancreaticoduodenectomy, the wide extent of the resection itself brings about functional deterioration of the digestive system, including endocrine insufficiency. In our study, incidence of postoperative diabetes was only 14.6%, which can be compared with other studies in the literature.13,14 Despite the obvious reduction in insulin secretion, patients undergoing pancreatic resection for periampullary malignancy usually do not have overt diabetes and actually show a notable improvement of glucose metabolism.15-17 Based on studies on the relationship between diabetes and pancreatic cancer, it was hypothesized that profound peripheral insulin resistance caused by a tumor-associated diabetogenic factor may be responsible for the high frequency of diabetes in patients with pancreatic cancer.15-17 In our study, we observed one case of a total disappearance of diabetes after pancreatic cephalic resection for pancreatic adenocarcinoma. Operative morbidity and mortality after duodenopancreatectomy is decreasing with improvements in operative techniques and intensive care methods. Operative mortality is still about 5% to 8%.18,19 The leading cause of this high mortality is pancreatic fistula. Rate of pancreatic fistula after pancreaticoduodenectomy has been reported to be 5% to 19%.20,21 In an attempt to reduce the occurrence of pancreatic fistula, many modifications in the management of the remnant pancreatic stump have been developed. The standard Whipple procedure proposed a pancreaticojejunostomy and later pancreaticogastrostomy was reported. Many reports favoring pancreaticogastrostomy
Rault et al
Pancreaticojejunal Anastomosis
243
have appeared recently. Kim and colleagues22 studied 86 patients to compare pancreatic leakage after pancreaticoduodenectomy between PJA and PGA. They reported that pancreatic leakage occurred at rates of 15.8% and 2.1%, respectively. Fabre and colleagues23 studied 160 patients who underwent PGA. They reported that pancreatic leakage occurred in 2.5% of the patients and that postoperative mortality rate was 3%. They also reviewed the large amount of literature concerning PGA and reported that the rate of pancreatic leakage was 4% on average.22,23 In a prospective randomized trial of about 145 patients, the John Hopkins group2 reported no difference in leakage rate between PJA and PGA (11.1% versus 12.3%). In our study, no statistically significant difference was found between PJA and PGA. We also observed that no marked difference was noted between patients with or without histologic obstructive pancreatitis, regardless of type of anastomosis. In the John Hopkins hospital experience,2 in a univariate analysis, pancreatic soft tissue appeared to be a factor considerably increasing risk of pancreatic fistula. In summary, with a median followup of 75 months after operation, patients who underwent pancreaticoduodenectomy with PGA presented considerably more pancreatic exocrine insufficiency (p ⬍ 0.025) than patients with PJA if there were no histologic obstructive pancreatitis. Complication rates of these two procedures are not statistically different. We recommend, each time it is possible, to perform pancreatectomy with PJA anastomosis. REFERENCES 1. Whipple AO, Parsons WB, Mullins CR. Treatment of pancreatic carcinoma of the ampulla of Vater. Ann Surg 1935;102: 763–779. 2. Yeo CJ, Cameron JL, Maher MM, et al. A prospective randomized trial of pancreaticogastrostomy versus pancreaticojejunostomy after pancreaticoduodenectomy. Ann Surg 1995;222: 580–588. 3. Alberti KGMM, Hockaday TAR. Diabetes mellitus. In: Weatherall DJ, Ledingham JGG, Warell DA, eds. Oxford textbook of medicine. 2nd ed. New York: Oxford Medical Publications; 1987:51–101. 4. Shinchi H, Takao S, Fukura K, Aikou T. Gastric acid secretion and gastroduodenal pH after pancreaticogastrostomy in dogs. J Surg Res 1996;61:152–158. 5. Takada T, Yasuda H, Uchiyama K, et al. Pancreatic enzyme activity after a pylorus-preserving pancreaticoduodenectomy reconstructed with pancreaticogastrostomy. Pancreas 1995;11: 276–282.
244
Rault et al
Pancreaticojejunal Anastomosis
6. Andersen HB, Baden H, Brahe NE, Burcharth F. Pancreaticoduodenectomy for periampullary adenocarcinoma. J Am Coll Surg 1994;179:545–552. 7. Jang JY, Kim SW, Park SJ, Park YH. Comparison of the functional outcome after pylorus-preserving pancreatoduodenectomy: pancreatogastrostomy and pancreatojejunostomy. World J Surg 2002;26:366–371. Epub 2002 Jan 15. 8. Lankisch PG, Droge M, Hofses S, et al. Steatorrhoea: you cannot trust your eyes when it comes to diagnosis. Lancet 1996; 347(9015):1620–1621. 9. Nakamura T, Tando Y, Terada A, et al. Can pancreatic steatorrhea be diagnosed without chemical analysis? Int J Pancreatol 1997;22:121–125. 10. Loser C, Mollgaard A, Folsch UR. Faecal elastase 1: a novel, highly sensitive, and specific tubeless pancreatic function test. Gut 1996;39:580–586. 11. Stein J, Schoonbroodt D, Jung M, et al. Measurement of fecal elastase 1 by immunoreactivity: a new indirect test of the pancreatic function. Gastroenterol Clin Biol 1996;20:424–429. 12. Katschinski M, Schirra J, Bross A, et al. Duodenal secretion and fecal excretion of pancreatic elastase-1 in healthy humans and patients with chronic pancreatitis. Pancreas 1997;15:191–200. 13. Lemaire E, O’Toole D, Sauvanet A, et al. Functional and morphological changes in the pancreatic remnant following pancreaticoduodenectomy with pancreaticogastric anastomosis. Br J Surg 2000;87:434–438. 14. Van Berge Henegouwen MI, Moojen TM, van Gulik TM, et al. Postoperative weight gain after standard Whipple’s procedure versus pylorus-preserving pancreatoduodenectomy: the influence of tumour status. Br J Surg 1998;85:922–926.
J Am Coll Surg
15. Ahren B, Andren-Sandberg A. Capacity to secrete islet hormones after subtotal pancreatectomy for pancreatic cancer. Eur J Surg 1993;159:223–237. 16. Permert J, Ihse I, Jorfeldt L, et al. Improved glucose metabolism after subtotal pancreatectomy for pancreatic cancer. Br J Surg 1993;80:1047–1050. 17. Permert J, Adrian TE, Jacobsson P, et al. Is profound peripheral insulin resistance in patients with pancreatic cancer caused by a tumor-associated factor? Am J Surg 1993;165:61–66; discussion 66–67. 18. Baumel H, Huguier M, Manderscheid JC, et al. Results of resection for cancer of the exocrine pancreas: a study from the French Association of Surgery. Br J Surg 1994;81:102–107. 19. Trede M, Schwall G, Saeger HD. Survival after pancreatoduodenectomy. 118 consecutive resections without an operative mortality. Ann Surg 1990;211:447–458. 20. Wade TP, el-Ghazzawy AG, Virgo KS, Johnson FE. The Whipple resection for cancer in U.S. Department of Veterans Affairs Hospitals. Ann Surg 1995;221:241–248. 21. Cameron JL, Pitt HA, Yeo CJ, et al. One hundred and forty-five consecutive pancreaticoduodenectomies without mortality. Ann Surg 1993;217:430–435; discussion 435–438. 22. Kim SW, Youk EG, Park YH. Comparison of pancreatogastrostomy and pancreatojejunostomy after pancreatoduodenectomy performed by one surgeon. World J Surg 1997;21: 640–643. 23. Fabre JM, Arnaud JP, Navarro F, et al. Results of pancreatogastrostomy after pancreatoduodenectomy in 160 consecutive patients. Br J Surg 1998;85:751–754.