Pancreaticojejunostomy (PJ) after pancreatico-duodenectomy (Duct to mucosa versus invaginating end to side anastomosis)

Pancreaticojejunostomy (PJ) after pancreatico-duodenectomy

(Duct to mucosa versus invaginating end to side anastomosis)

Mahmoud Elganzoury,MD; MRCS; Mostafa Fouad, MD;

Mohammed Abdo,MD; MRCS; Amr kame/,MD

Department of Surgery, Ain Shams University, Cairo, Egypt.

Abstract

Aim: This prospective randomized trial compares the results of pancreaticoduodenectomy with duct to mucosa pancreaticojejunostomy with external stent versus pancreatico-duodenectomy with invaginating (dunking) end to side pancreaticojejunostomy

Methods: This study included 40 patients who underwent surgical therapy for cancer head of  pancreas, between Sep.  2009  and  March 2011  at  Ain  Shams University Hospitals.

Patients were randomly assigned to surgical procedures. Patients in group 1 (n 20) underwent pancreaticoduodenectomy with end to side duct to mucosa pancreatico-jejunostomy and external stent, and patients in group 2 (n 20) underwent pancreatico-duodenectomy with invaginating (dunking) end to side pancreaticojejunostomy.

Results: In group 1 two patients (10%) developed pancreatic fistula compared to five patients (25%) in group 2.As regards development of intra-abdominal abscess, one patient in group 1 developed  that complication  while two patients (10%) in group 2 developed intra-abdominal abscess. One patient (5%) in group 2 had postoperative bleeding.One patient in group 1 (5%) developed  septicemia  compared to three patients  (15%) in group 2. Two patients in group 2 needed reoperation.

Conclusion: Duct to mucosa pancreaticojejunostomy has less post-operative  morbidity as pancreatic fistula, intra-abdominal abscess  and  septcemia than  invaginating (dunking) pancreaticojejunostomy.

Introduction:

Whipple and colleagues  (1935) in New York ligated the main pancreatic duct as part of their two-stage pancreaticoduodenectomy

for  cancer  of  the  periampullary region.I

Mortality after pancreaticoduodenectomy has dropped significantly  over the last few decades from 25% to less than 5%. This is due to  better patient selection , high-dependency care postoperatively, and centralization  of patients to  regional  surgical  centers.2

The pancreaticojejunostomy anastomosis is   nicknamed  the   11Achilles  heel 11      of panceaticoduodenectomy because of the potentially disastrous sequalea of life­ threatening intra-abdominal sepsis and hemorrhage.3

Pancreatic fistulas remains a major cause of postoperative complications; it is reported that the incidence of pancreatic fistulas after PD is 6%-25%.4

Roder et al. (1999) showed that external drainage of the pancreatic duct decreased the rate of pancreatic fistula from 29.3% to 6.8%, and reduced the median hospital stay from 29 to 13 days.s

Li-Ling and Irving (2001)  approved  that prophylactic administration of somatostatin does not reduce the incidence of pancreatic anastomotic leak,  overall morbidity, or mortality  after  pancreatic  resection.6

Patients and methods:

This prospective study included 40 patients who underwent surgical therapy for cancer head of pancreas, between Sep. 2009 and March 2011at Ain Shams University Hospitals. Patients were randomly assigned to surgical procedures. Patients in group  1 (n 20) underwent pancreaticoduodenectomy with end to side (duct to mucosa) pancreaticojejunostomy and external stent, and

patients in group 2 (n 20)  underwent pancreaticoduodenectomy with invaginating (dunking) end to side pancreaticojejunostomy.

Patients were diagnosed as cancer head of pancreas with  Multidetector CT  scan  of abdomen and elevated CA 19-9.

Patients having distant metastasis, locally

advanced tumors with invasion of portal or superior mesenteric veins were excluded from the present study.

We have used the International Study Group

for Pancreatic Fistula (ISGPF) guidelines defining pancreatic fistula as any measurable volume of fluid on or after postoperative day

3 with amylase content greater than 3 times

the serum amylase activity.3

Operative Teehnique:

End  to    side  (duct  to  mucosa)

pancreati.cojejunostomy:

The outer layer is performed with posterior interrupted stitches of 4/0  prolene  which extends from the pancreatic capsule to the jejunal mucosa Figure(l).A 2-3 mm opening is  made  in  the  jejunum  adjacent to  the pancreatic duct The anastomosis between duct and mucosa of jejunum is performed with 5/0 prolene intrrupted stitches with an external drainage stent ( 6 french epidural catheter) placed  across the PJ anastomosis  into the pancreatic duct and brought out externally via the jejlm.alloop and abdominal wall Figure(l). The  outer  layer  is  completed anterior.

Figure (1): Duct to mucosa PJ: Posteriorouterrow is complete, right angle in the pancreatic duct.

Figure (2):Duct to mucosa PJ:  Pancteaticstent is brought out through jujenum.

End  to  side  invaginating (dunking)

pancreaticojejunostomy:

A row  of  interrupted sutures is placed through the dorsal capsule, with the needle entering the pancreatic tissues at about 1 em away from the cut end of the stump and taking an adequate bite of the parenchyma.The needle is driven through the serosa of the jejunal limb at about 1 to 2 em from the mesenteric border. A longitudinal enterotomy is made about 5 to

10 mm away from the posterior row of sutures. The  posterior inner  row of sutures is then constructed, first catching the pancreas in an inside-out fashion, ensuring an adequate bite of the parenchyma, and exiting just at the cut edge of the remnant  pancreas. The suture is driven through the  posterior edge  of  the enterotomy, ensuring a full-thickness bite, in an outside-in fashion. All knots are tied after completion of the suture row. This is followed by the anterior inner row of sutures. The fmal step is the second row of anterior sutures, the aim of which is to cover the inner suture line by  imbricating the  jejunal serosa  over  it.

Surgical drains were placed near not at the pancreaticojejunostomy  and the choledochojejunostomy (One learns early in training not  to  place  a drain  directly on ananastomosis because it encourages leakage) and were connected  to a closed drainage system. Pancreatic tube  was  connected to urinary bag and was removed after 6 weeks if there is no pancreatic fistula.

We always  use loupes  for magnification

(x 3).

No prophylactic somatostatin or Octreotide was used.

Statistical analysis:

Analysis of data was done by ffiM computer using  SPSS  (statistical program for  social science version 12) as follows:

• Description of quantitative variables as mean, SDandrange

• Description of qualitative variables as number and percentage

* Chi-square test  was  used   to  compare

qualitative variables between groups.

* Fisher exact test was used instead  of chi­

square when one expected cell less than or

equal to 5.

* Unpaired t-test was  used   to  compare

quantitative variables,  in parametric data

(SD<50% mean).7

P value >0.05 insignificant

P<0.05 significant

P<0.01 highly  significant.

Results:

During the period between September 2009 and  March  2011,  40 patients  diagnosed as cancer head of  pancreas underwent pancreaticoduodenectomy. Group 1 included

20 patients  who underwent pancreaticoduodenectomy with end to side duct to mucosa pancreaticojejunostomy and external stent, and patients in group 2 (n 20) underwent pancreaticoduodenectomy with invaginating (dunking) pancreatico­ jejunostomy.

Group 1 included 15 males and 5 females;

the mean age was 62.4±2.8. Group 2 included

13 males  and 7 females;  the mean age was

61.2±1.3 Table(l).

Table(1):Comparison between both studied groups as regard general data.

This table shows that no statistically significant difference couldbe detected between both   groups  as  regard  general  data.

In group 1 two patients (10%) developed pancreatic fistula compared to five patients (25%) in group 2. As regards development of intra-abdominal abscess, one patient in group

1 developed  that  complication while  two patients (10%) in group 2 developed intra­ abdominal abscess. One patient in group 1 (5%) developed septicemia compared to three patients (15%) in group 2.

In group 1, one patient of the 2 patients

with pancreatic fistula required percutaneous drainage of an infected  intra-abdominal collection and he was treated with somatostatin, total parenteral nutrition and antibiotics for 22 days. Sepsis was relieved and the patient improved. The  other  patient  was  treated conservatively with somatostatin, oral nutrition and antibiotics for 19 days.

In group 2, two patient of the five patients

with pancreatic fistula required percutaneous

drainage of an infected intra-abdominal collection and they  were treated with somatostatin, total parenteral nutrition and antibiotics. One of these two patients underwent reoperation on the 11th day postoperative due to peripancreatic collection & sepsis. Drainage of collection and peritoneal toilet were done, there was no evidence of pancreatic leakage. Two tube drains were used for drainage. One patient had clinical leakage with fever and leukocytosis, he developed intra-abdominal bleeding, emergency exploration was done. We found bleeding from the stump of a gastroduodenal artery (because of erosion by local   sepsis   and  the  pancreatic juice's proteolytic enzymes). We ligated the stump of a gastroduodenal artery and peritoneal toilet was done. Two tube drains were used for drainage.Postoperative this patient was treated successfully with somatostatin, total parenteral nutrition and antibiotics.The other two patients were treated conservatively with somatostatin, oral nutrition and antibiotics.

Table (2): Comparison between both studied groups as regard complications.

Table (3): Comparison between both studied groups as regard morllllity.

Table(2) shows that complications were more frequent among dunking, statistically significant difference couldbe detected between both   groups  as  regard  general  data.

As regards the mortality, there was one case (5%)  of mortality in group1 due  to bronchopneumonia compared to two (10%)

mortality cases in group 2, one patient developed pancreatic leakage and died of septic shock and the other patient died due to pulmonary embolism Table(3) and Figure(3).

Table(3) shows no statistically significant

difference  between both groups as regard mortality  by  using Fisher exact test.

lmDuct to mucosa mDunking

%

100

80

60

40

20

0

No                                                   Yes

Figure (3): Comparison between both studied group as regard mortality.

The average hospital stay in group 1 was

12.6±3.4 days compared to 27.7±6 days in

group  2 and according to these  results pancreaticoduodenectomy with end to side duct to mucosa pancreaticojejunostomy and external stent has lower median hospital stay

compared to pancreaticoduodenectomy with invaginating (dunking) pancreaticojejunostomy with statistically highly significant difference in between by using unpaired t-test Table(4) and Figure(4).

Table (4): Comparison between both studied groups as reglll'd hospital stay.

This table shows that dunking group had higher median hospitals stay compared to duct to mucosa  group  with statistically highly

Median duration

30

25

20

15

10

5

0

Duct to mucosa

significant  difference  in between by using unpaired t-test.

Dunking

Figure(4):Comparisonbetweenboth studied groups as regard hospital stay.

Discussion:

Pancreatic fistula,  leak, and abscess  are major contributors to the morbidity of pancreatic resection. The reported prevalence of these complications varies widely  in the literatures. This  variation is  secondary to differences in definition and to the retrospective nature of most reports. To improve mortality and  morbidity, the  rate  of  occurrence of pancreatic fistulas after pancreaticojejunostomy after pancreatic resection  must be reduced.

Our study showed that   pancreatic fistula was more frequent among dunking  PJ cases (25%) than patients who  underwent pancreaticoduodenectomy with  end to side duct to mucosa pancreaticojejunostomy (10%). Statistically significant difference could be detected between both  groups (P  0.05).

Yeo et al., reported  that pancreatic fistulas were correlated with anastomotic  technique, operative time, a surgeon's skills and experience in performing a PD, tumor location, and co-morbid illnesses.

In study performed by Schmit et al., a duct­

to-mucosa anastomosis was performed in 453 patients, and an invaginated anastomosis was performed in 52 patients. In the remaining  5 patients, 3 underwent Peustow reconstruction of the remnant, and in 2, the reconstruction was  unclear  based  upon  operative records. There was a more than 3X higher incidence of    PF   in  invaginated  anastomoses.

Bartoli et al., observed a higher incidence of fistulas  after end-to-side PJ anastomosis (16.5%) than after end-to-end (11.7%) or duct­ to-mucosa (11.5%) anastomosis.

Between Aprill996 and March 2006, Choe et al., studied  172 consecutive patients who had undergone PD.  The methods used  for anastomosis were  divided into  end-to-end dunking anastomosis between cross-sections of the jejunum and the pancreatic stump Group(!), and  pancreatic duct  to  jejunal mucosal  anastomosis  Group(2). Leakage  of the pancreatico-jejunostomy occurred in 30 of

172 patients (17.4%). Incidence of pancreatic fistula in group  1 is 16.5%  & in group 2 is

19.6% ( p value 0.582).

Crist and Cameron observed no difference in the incidence of fistulas between invagination (13%) and duct to mucosa (16%) anastomosis.

Similar data are reported by Grace et al., with fistula rates of 6% to 13% after invagination anastomosis and 11% to 14% after duct-to­ mucosa anastomosis.

Roder  et al., reported a prospective but nonrandomized study in 85 patients showing that  external drainage of  pancreatic duct decreased the pancreatic fistula rate from 29.3% to 6.8% compared with no stent.

Tani  et  al.,  reported the  incidence of pancreatic fistula according to the International Study Group on Pancreatic Fistula criteria was not different (external, 20%; vs internal, 26%), and the incidence  of the other complications was similar between stent types. The median postoperative hospital stay was 21 days (range,

8-163 d) in the internal drainage group, which was shorter than the median stay of 24 days (range, 21-88 d) in the external drainage group (P  0.016).

In conclusion, this prospective randomized

study showed that duct to  mucosa pancreaticojejunostomy has less post-operative morbidity as pancreatic fistula, intra-abdominal abscess, septcemia thaninvaginating (dunking) pancreaticojejunostomy. But it is not affecting the mortality.

References:

1- Whipple AO, Parsons WB, Mullins CR: Treatment of carcinoma of the ampulla of vater. Ann  Surg  1935; 102:  763-779.

2- Yeo CJ, Cameron JL, Sohn TA: Six hundred fifty consecutive  pancreatico­ duodenectomies in the 1990s. Ann Surg

1997; 226: 248-260.

3- Bassi C, Dervenis C, Butturini G: "Postoperative pancreatic fistula: An international study group (ISGPF) definition," Surgery 2005; 138(1): 8-13.

4- Strasberg SM, Drebin JA, Soper  NJ: Evolution and current status of the Whipple procedure:      An      update   for gastroenterologists. Gastroenterology1997;

113: 983-994.

5- Roder  JD, Stein  HJ, Bottcher KA, et al: Stented        versus   nonstented pancre aticoj ej uno stomy   after pancreatoduodenectomy. A prospective study. Ann  Surg   1999; 229:   41-48.