A Comparative Study between Duct-to-Mucosa and Invagination Technique for Reconstruction after Pancreaticoduodenectomy: Aprospective Study

A Comparative Study between Duct-to-Mucosa and Invagination Technique for Reconstruction after Pancreaticoduodenectomy: Aprospective Study

Mohamed El Serafe, MD; Ahmed Hussein Abd El hafiz, MD; Mohamed Shaaban Khalifa, MD.

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

Background: The pancreaticojejunostomy has notoriously been known to carry a high rate of operative complications,morbidity,and mortality mainly due to anastomotic leak and ensuing septic complications.

Patients and methods: From January 2012 to October 2015, we presented a prospective study which included 24 patients who underwent pancreaticoduodenectomy (PD) operation through either Whipple resection or modified Whipple(pylorus-preserving).Patients were reviewed and divided into 2 groups (A,B) according to the type of pancreaticojejunostomy (PJ), (invagination vs duct-to-mucosa ).

Results: 12 patients in each group were operated on: Group A patients had invagination technique for PJ ,while all the 12 patients in group B had duct –to-mucosa anastomotic technique for PJ. 1 (8.3%)case in group A developed pancreatic fistula (PF),while 3 (25%) cases in group B developed PF, and 1 case (8.3%) in group A had mild anastomotic leak which was managed conservatively,while 3 cases (25%) in group B developed moderate to severe anastomotic leak with intra-abdominal collection which required CT-guided percutaneous drainage and operative intervention. Average age was (mean ±SD) = (55 ±12),average operative time was (245 ±75) min.

Conclusion: PF after PD represents a critical trigger of potentially life-threatening complications. Although the best method for dealing with the pancreatic stump after PD remains controversial, recent reports described the invagination technique todecrease the rate of PF significantly compared to the duct-to-mucosa technique. Our results appeared to be closely related to the published literature.

Key words: Pancreatic fistula (PF), pancreaticodudenectomy (PD), pancreaticojejunostomy

(PJ), invagination technique, duct-to-mucosa technique.

Introduction:

Theindicationsofpancreaticodudenectomy (PD) have expanded to encompass a broad spectrum of periampullary tumors including both benign and malignant lesions, chronic pancreatitis, and occasionally trauma. During the last decade, although the rate of operative mortality  significantly decreased  after  PD, the  incidence  of  post-operative  morbidity still remains high..1,2

The occurance of PF is a critical trigger of  life-threatening  complications  such  as intra-abdominal abscess and hemorrhage,3 whichis also potentially associated with markedly prolonged hospitalization. Most of the large PD series have reported rates of PF of over 10%.4,5 Risk factors for PF depends upon:  1) general  patient  factors  including age,6 sex,7 DM,8 and nutrition;9 2) disease- related factors,including pancreatic duct size,pancreatic     texture,and     pathology;10 3) procedure-related factors including, blood loss,     operative   time,11    and   anastomotic technique.12 Among the risk factors, the most important might be the texture of the remnant pancreas.13

Indeed, despite an occurance rate of PF of  5%  in  cases  of  hard  pancreatic  tissue, the rate rises to nearby 20% in cases of soft pancreatic tissue.14 The risk of developing a PF is significantly associated with the final histopathological diagnosis of the resected specimen, with lower risk in adenocarcinoma, and high risk in cystic neoplasm, or diseases originating  from  the  bile  duct.15   This  is also because pancreatic malignancy usually causes  main  pancreatic  duct  dilatation which also occurs in chronic pancreatitis; therefore, a fibrotic hard remnant pancreas and an enlarged duct are easily anastomosed, whereas soft pancreas remains at risk of developing pancreatic fistula due to its fragility and its secretion of large amounts of pancreatic juice.16  Surgical technique might be one improvable aspect of PD that can reduce the pancreatic leakage rate;it is critical in the management of the pancreatic remnant  because  of  the  various  methods used by surgeons. Methods of reconstruction which are used between pancreatic stump and intestine include end-to-side with or without duct-to-mucosa anastomosis, end-to-end invagination technique.17

Patients and methods:

From  January  2012  to  October  2015, we reviewed 24 consecutive patients,who underwent pancreaticodudenectomy (PD), all patients were studied and evaluated. Patient demographic data,full history and clinical examination, full preoperative laboratory and radiological  investigations  were  obtained. The patients had either a classic Whipple resection (n=16) (66.7%), or pylorus- preserving PD (n=8) (33.3%), according to the decision of the attending surgeon. All patients  were  thoroughly  informed  about the procedure,and signed consent was obtained. Preoperative evaluation consisted of an extensive imaging workup including computed tomography (CT), magnetic resonance imaging (MRI), and magnetic resonance cholangio-pancreatography (MRCP),  to  rule  out  metastatic  disease  or encasement  of  superior  mesenteric  vessels (SMV).  Transduodenal  fine  needle  biopsy (FNA), and endoscopic biliary stenting were occasionally  used  whenever  required,  the final assessment of the respectability of the malignant tumor was made intraoperatively, based upon findings such as liver metastases not identified preoperatively and involvement or encasement of SMV or portal vein. All patients    received                       preoperative          antibiotic prophylaxis,        and  low     molecular         weight heparin   was   administered   for   4   weeks. None  of  the  patients  received  octreotide analogue  preoperatively  or  postoperatively. The pancreatico-jejunostomy            anastomotic technique used was either (A) invagination or (B) duct-to-mucosa,based on these variables

,patients were divided into 2 groups, group A had invagination technique, and group B had duct-to-mucosa anastomotic technique. Epidemiological, operative, and histological data are presented in Table (1).

Surgical technique:

After  completion  of  the  standard  PD and thorough hemostasis, the jejunum was transected  at  about  20  cm  from  ligament of   Treitzʼs,   reflected   upwards   through an  incision  in  the  transverse  mesocolon, and anastomosed end-to-side with the choledochus by one layer of interrupted 4-0 polydioxanone surgical sutures. At 20 cm distal to the biliary anastomosis, the jejunum was interrupted, and the end of the pancreas inserted into the bowel by means of either one of the following 2 techniques, regardless of the diameter of the pancreatic duct, (a) the invagination technique created with 2 rows of continuous or interrupted sutures done via Prolene 3-0 (Ethicon), between the cut edges of the jejunum (seromuscular layer) and the pancreatic parenchyma, directing the end of the pancreas to invaginate into the jejunum, and   (b)   the   end-to-side   duct-to-mucosa PJ  established  with  the  use  of  interrupted 5-0 or 6-0 PDS (Ethicon) sutures between the pancreatic duct and the mucosa of the jejunum, and reinforced by seromuscular sutures  from  the  jejunum  to  the  cutting edge of the pancreatic parenchyma. At 20 cm  distal  to  the  pancreatico-jejunostomy (PJ), the jejunum was anastomosed to the gastric pouch in an end-to-side fashion (gastrojejunostomy),   or   to   the   duodenal loop in case of pylorus-preserving resection (duodenojejunostomy). The alimentary tract reconstruction was accomplished via Roux- en-Y enteroenterostomy, finally a closed suction tube drain was placed near the PJ anastomosis and brought out through the right lateral abdominal wall Figures (1,2,3).18

Perioperative management-Data collection: All          patients            received           a  standard perioperative  management.  Details  of surgical procedure and complications were recorded,drain fluid volume and amylase levels were monitored daily, nasogastric (Ryle) tube was removed, and feeding was introduced as soon as adequate gastric emptying   was   re-established.   The   drain tube was removed 6 to 10 days post operatively,provided that amylase levels in the drainage fluid did not exceed serum levels by more than 3 times. When intra-abdominal collections were suspected, an abdominal CT was done and percutaneous aspiration was carried out for fluid culture and biochemical analysis. A percutaneous tube for drainage under CT guidance was inserted when pancreatic collection reoccurred or an intra- abdominal abscess developed.

A pancreatic fistula (PF) as a result of pancreatic leakage was defined as a drainage volume of more than 50 ml after the 8th postoperative day,with an amylase level greater than 1000 IU/ml, or greater than 3 times the serum amylase level. In accordance to the definition of the International Study Group of  Pancreatic  Fistula.19  When  a  diagnosis of a pancreatic fistula was established, the tube  remained  in  place  until  the  fistula volume was reduced to a level less than 50 ml/day. The severity of pancreatic leakage was graded as: (1) grade A, when it subsides without intervention; (2) grade B, when a non surgical intervention was applied to resolve the complication by using percutaneous CT- guided drainage, and appropriate antibiotic therapy;  and  (3) grade  C,  when  operative intervention was required.20

Results:

In this prospective study, 24 patients presented by pancreatic lesions in the form of malignant tumors, or benign conditions such as chronic pancreatitis. All the patients were subjected to pancreaticodudenectomy (PD) via Whipple resection or pylorus-preserving operation. All the patients had full laboratory and radiological investigations to assess operability ,and metastatic disease.Whipple resection was performed in 18 patients (75%), while pylorus-preserving procedure was performed in 6 patients (25%), according to the procedure chosen for PJ anastomosis,the 24 patients were classified into 2 groups (A,B). Group A involved 12 patients underwent PJ anastomosis via invagination technique,while group B (12) patients underwent PJ anastomosis via duct-to-mucosa technique,average  age  of  the  patients  was (55 ±12) years, average operative time was (245 ±75) min,with an average intraoperative blood loss (750±200cc), the average hospital stay (12 ±2.5) days. Total parenteral nutrition (TPN) was instituted in those patients who developed any anastomotic defect and delayed gastric emptying. Both groups of patients were comparable regarding epidemiological, histological, and intraoperative data Table (1).

As regard pancreatic leakage incidence,1 patient (8.3%) in group A developed pancreatic anastomotic leakage with mild intra-abdominal collection which subsided conservatively, while 3 patients (25%) in group B developed moderate to severe pancreatic leakage, pancreatic leakage incidence in relation to the type of PJ anastomosis is presented in Table (2). The use of invagination technique for PJ anastomosis was associated with a significantly lower pancreatic leakage rate as compared with the use of duct-to-mucosa technique (8.3% vs 25%, P =0.035).

Postoperative complications, morbidity, hospitalization, and mortality are presented in  Table (3).  3  patients  (75%)  in  group (B) developed PF, 2 of them had grade B pancreatic fistula which required CT-guided percutaneous drainage and antibiotic therapy, the 3rd patient had grade C pancreatic fistula which required surgical intervention, the disruption of the PJ anastomosis that was found in this patient was managed by means of stable closure of the jejunal loop, and external drainage of the pancreatic stump. While in group (A), 1 patient only (8.3%) developed grade A pancreatic fistula, which resolved and closed spontaneously.

The overall morbidity was significantly high in group (B) patients who had duct-to- mucosa than those of group (A) who had invagination  technique  for  PJ  anastomosis (the overall morbidity rate in group A was 33.3%  vs  83.3%  in  group  B,  P=0.0046). The length of hospital stay was significantly increased in group (B) patients, than those of group (A) patients (P <0.01). The overall mortality rate in our study was 2.8%. Two patients (16.6%) died in group (B) due to major anastomotic leak and severe intraperitoneal sepsis, while 1 patient (8.3%) in group (A) died due to massive myocardial infarction. 3 patients (25%) in group (B) had developed moderate to severe anastomotic dehiscence and leakage, 2 of them were managed via CT- guided percutaneous drainage and antibiotic coverage,and the 3rd one needed surgical intervention, and open drainage of intra- abdominal abscess. While 1 patient (8.3%) in group (A) developed minor anastomotic leak and mild collection which was managed conservatively.

Discussion:

Drainage of the pancreatic remnant to the gastrointestinal tract has been thoroughly investigated, and various techniques have been reported. Despite sporadic reports describing low rates of pancreatic leakage (from 3% up to 14.3%) after pancreaticogastrostomy, no valid conclusion can be drawn regarding its superiority against PJ, and further randomized controlled trials are required.21

PJ is still the most commonly used method of restoring pancreatico-enteric continuity after PD, and its technical improvements are essential to reduce the pancreatic leakage rate.22

Regarding  anastomotic  techniques  used for  reconstruction  between  the  pancreatic cut  surface  and  the  jejunum,  both  end-to- side  duct-to-mucosa  anastomosis  or  end- to-end invagination techniques have been extensively investigated.  Certain  reports have shown no clear evidence for or against one particular method of pancreatico-enteric anastomosis.23

The choice of pancreatic anastomosis method might be based on individual experience and adherence to basic principles such  as  good  exposure  and  visualization, fine, non strangulating suture placement, to produce  a  patent,  watertight  anastomosis; and preservation of blood supply. Recent reports  described  the  invagination  method to decrease the rate of pancreatic fistula significantly compared to the duct-to-mucosa anastomosis.24

In our study, the incidence of PF in group (B) patients operated upon via duct-to-mucosa technique, and the incidence of anastomotic dehiscence and leakagewas relatively higher as compared to group A patients who were operated upon via invagination technique, 3 patients (25%) in group (B) developed PF, compared to 1 patient (8.3%) in group (A) developed PF, also 3 patients (25%) in group B  developed  severe  anastomotic  leakage and managed via CT-guided percutaneous drainage and operative intervention, in comparison to 1 patient (8.3%) in group A developed mild anastomotic dehiscence and mild collection managed conservatively.

Mortality rate was higher in group B (16.6%) (n=2) as compared to group A (8.3%) (n=1), in our study TPN was given to all patients who developed PF and anastomotic leakage.

Ferguson     and     Wangesteen     (1998); were the 1st to report a new interesting experimental technique for PJ anastomosis, directly approximating the jejunal mucosa to the Wirsung duct epithelium, subsequently described by Madden.25 Since then, other modifications have been suggested, such as the 2 layers technique, the use of fibrin glue, and the use of ultrasonic shears.26

Weagnee with Sikora and Posner (1995),

Figure (1): Hepaticojejunostomy.                          Figure (2): Enteroenterostomy.

Figure (3): Gastrojejunostomy.

Marcus et al;and Suzuki et al (2002); using the selective duct-to-mucosa technique when the Wirsung  duct  is  dilated  (>5  mm)  and in the presence of firm fibrotic pancreas,27 whereas the invagination technique appears to be safer and then should be performed in the presence of friable pancreas with non- dilated duct, considering the invagination technique for PJ is always feasible in all situations (small or dilated duct, friable or

Table (1): Epidemiological, operative, and histological data of patients.

*Mean±SD

Table (2): Comparison of the pancreatic leakage rates in the 2 groups.

firm, fibrotic pancreas),28 it should be the procedure of choice. The duct-to-mucosa technique  after  PD  in  case  of  a  normal, non-dilated Wirsung duct is a more time consuming, and demanding technique, necessarily requiring a microsurgical skills, but it can be alternatively adopted in case of an enlarged gland that does not fit well in the jejunum to be invaginated.29,30

Conclusion:

Although  the  best  method  for  dealing with the pancreatic stump after PD remains controversial, recent reports described that the invagination technique to decrease the rate  of  PF  significantly  compared  to  the duct-to-mucosa technique, although the low number of patients in our series that did not allow definitive conclusion, more time, and

Table (3): Postoperative complications, and outcome of patients.

experience  is  needed,  for  more  accurate results.31,32

Our results encouraged the use of invagination technique as compared to the duct-to-mucosa technique, as it seems to be more save, feasible, and associated with less incidence   of   postoperative   complications, less morbidity, mortality,and shorter hospital stay.

Reference:

1-    Buchler  MW,  Wagner  M,  Schmied  BM: Changes       in   morbidity   after   pancreatic resection; toward the end of completion pancreatectomy.   Arch   Surg   2003;   138:1310–1315.

2-   Tsuchiya R, Tsunada T, IschidaT: Resection for cancer of the pancreas. The Japanese experience.  Baillieres  Clin  Gastroenterol2012; 4: 931–939.

3-   Neoptolemos JP, Russel RC, BramhallS: Low mortality following resection for pancreatic and periampullary tumors in 1016 patients: UK survey of specialist pancreatic units. UK Pancreatic Cancer Group. Br J Surg 2008;84: 1370–1376.

4-   Kleespies  A,  Albertsmeir  M,  Obeidat  F: The challenge of pancreatic anastomosis. Langenbecks Arch Surg 2008; 39: 450–456.

5-   Bottger       TC,       JungingerT:       Factors influencing morbidity and mortality after pancreaticoduodenectomy; critical analysis of 221 resections. World J Surg 2009; 23:164–171.

6-   Buchler MW, Friess H, Wagner M: Pancreatic fistula after pancreatic head resection. Br J Surg 2010; 87: 883–889.

7-   Gouma DJ, van Geenen RC, van Gulik YM: Rates of complications and deaths after pancreaticoduodenectomy; risk factors and the  impact  of  hospital  volume.  Ann  Surg2008; 232: 788–795.

8-   Sohn TA, Yeo CJ, Cameron JL: Resected adenocarcinoma of the pancreas-616 patients; results, outcomes, and prognostic indicators. J Gastroenterol Surg 2011; 4: 138–146.

9-   Wade  TP,   el-Ghazzawy  AG,  Virgo   KS: The Whipple resection for cancer in U.S. Department  of  Veterans  Affairs  Hospital. Ann Surg 2002; 221: 241–248.

10- Halloran   CM,   Ghareh   P,   BosonnetL: Complications of pancreatic cancer resection. Dig Surg 2002; 19: 138–146.

11- Yeh     TS,     Jan     YY,     Jeng     LB: Pancreaticojejunal anastomosis leak after pancreaticoduodenectomy;     multivariate analysis of perioperative risk factors. J Surg Res 2005; 67: 119–125.

12-  Crippa S, Salvia R, Falconi M: Anastomotic leakage  in  pancreatic  surgery.  Br  J  Surg 2007; 9: 10–15.

13-  Tran   K,   Van   Eijck   C,   Di   Carlo   V   : Occlusion of the pancreatic duct after pancreaticojejunostomy:   A     prospective randomized  trials.  Ann  Surg  2012;  236:422–428.

14-  Papardoriston   DN,    De     Agostine    H, Fortner  JG:  Ligation  of  pancreatic  ductin  pancreatectomy.  Br  J  Surg  2006;  67:260–265.