The five-year survival rate of pancreatic cancer is the lowest among all cancer types. Early diagnosis is essential and the only chance for cure is resection [1] [2]. Up to 85% of patients with pancreatic cancer are already in advanced stage at the time of the diagnosis and in those patients curative resection is possible in only 10–15 % [1]. Pancreaticoduodenectomy is the procedure of choice for all periampullary tumors and pancreatic head tumors, specifically [1] [3]. Although the mortality rate following PD is reported to be around 3–5% only [1], postoperative morbidity is still relatively high (20–30%) despite improvements over recent years [4]. Pancreaticoduodenectomy is an invasive and potentially risky procedure and a key step is anastomoses of the pancreatic remnant to the intestine. [5]. Various techniques for managing the pancreatic remnant have been reported including simple ligation of the pancreatic duct [6][7], occlusion of the pancreatic duct using a synthetic rubber injection or fibrin glue [8] [9], optimization of the blood supply of the edge of the pancreatic remnant and meticulous placement of sutures using magnification [10], application of fibrin glue sealant around the pancreaticojejunal (PJ) anastomosis [11], various modifications of the PJ (either end-to-end or end-to-side anastomosis) [12][13] or isolated Roux-en-Y pancreaticojejunostomy [14][15]. Results following PG have been inconsistently reported in numerous studies [16][17]. Bassi et al. showed that biliary fistula, postoperative collections and delayed gastric emptying were significantly reduced in patients treated by PG compared with alternative anastomoses [18] . In addition, PG is associated with lower frequency of multiple surgical complications. The purpose of the present study is to describe and evaluate our modified PG technique on 36 patients in a retrospective manner to determine whether this is a safe procedure for managing restoration of pancreatic remnant.

We conducted a retrospective review of 36 patients who underwent PG after PD at King Faisal Specialist Hospital & Research Center (Gen. Org.), Jeddah, KSA between 2008 and August 2015. All procedures were conducted by a single surgeon (Dr. Saud Al-Muhammadi (Saudi Arabia).

All patients were investigated preoperatively to confirm their diagnosis and to evaluate their general and nutritional health statistics. Blood work, electrocardiogram (ECG), Chest X-ray and an abdomen and pelvis computed tomography (CT) scan were performed for all patients.

Clinical and pathological data were obtained from the hospital data base and medical records. Data obtained for each patient included demographics, preoperative laboratory values, preoperative intervention, intraoperative data such as the use of prophylactic antibiotics, the use of intravenous octreotide before the anastomosis, the length of the operation and the blood loss.

Postoperative data included complications, pathological findings, and survival outcomes. Analyses were performed using statistical software (IBM SPSS statistics 21 for windows) with the t-test used where appropriate. Statistical significance was set at p = 0.05.

The abdominal cavity was accessed via a longitudinal upper midline incision and, after stepwise mobilization, the PD was performed. The pancreatic parenchymal transection was performed using cutting electrocautery after controlling the blood supply originating from two arterial arcades in front of and behind the pancreatic head using a 4-0 PDS suture (Video 1).

Reconstruction of the PG was done using the most common technique described by Mackie and Rhoads involving a simple invagination technique with a few modifications. A 2–2^-1/2 cm centimeters length of the remnant pancreas was mobilized from the surrounding tissues using ligature ties (almost always one to two splenic vein and arterial branches need to be ligated). Excessive mobilization of the pancreas was avoided to reduce the risk of pancreatic necrosis/pancreatitis.

A small incision was then made in the posterior surface of the stomach (the diameter of the gastric incision is calibrated to 1 cm less than the diameter of the pancreatic stump) and two invaginating sutures fixed to each corner of the gastrectomy incision. These two sutures are responsible for invaginating the pancreas into the stomach and are an important step during this procedure (Figure 1).

A row of interrupted 4-0 monofilament sutures is then placed between the gastric seromuscular layer and the anterior wall of the pancreatic parenchyma 2 cm from the cut edge of the pancreas. The suture is passed proximal to distal to improve invagination of 1–2 cm of pancreatic parenchyma when tying these sutures (Figure 2). Sutures are then placed from the posteroinferior gastric wall to the posterior body of the pancreas, using the same technique of proximal to distal passage of the suture through the pancreas (Figure 3).

A second continuous layer of sutures is then added to the posterior body of the pancreas and the inferior cut edge of the gastrostomy (Figure 4).

The hepaticojejunostomy and gastrojejunostomy are performed using a Roux-en-Y technique with an interrupted single layer of PDS for the H-J and a stapled anastomosis for both the gastrojejunostomy and the enteroenterostomy.

Two drains are left in place, one in the sub hepatic area and one below the pancreatic anastomosis. The abdominal cavity was closed using a running suture of loop #1 PDS. The skin was closed using a stapler. The median operation time was 6 hours, ranging between 2–11 hours. The median estimated blood loss 400 ml, ranging between 100–1500 ml.

Video 1: Pancreaticogastrostomy procedure.

All patients were transferred to the surgical intensive care unit (SICU) for 24 hours postoperatively. Prophylactic intravenous (IV) cefazolin (Kefazol®) 1 g every 8 hours and metronidazole (Flazol®) 500 mg every 6 hours was given for seven days, metoclopramide (Primperan®) 10 mg every 8 hours PRN for nausea and vomiting, omeprazole (Losec®) 20 mg every 12 hours, all our patients received either patient controlled analgesia (PCA Fentanyl) (Sublimaz®) or epidural anesthesia for 3 days maximum. Acetaminophen (Perfalgan®) 500–1000 mg every 6 hours and tramadol (Tramal®) 50–100 mg every 6 hours were used when the PCA or epidural anesthesia was removed, octreotide (Sandostatin®) 100 mcg every 8 hours was given for 5 days. Patients received unfractionated heparin 5000 IU TID via a subcutaneous (SC) route as DVT prophylaxis.

Oral fluids were reintroduced on day-2 and then progressed to full diet, as tolerated. Delayed gastric emptying was defined following the definition of the international study group for pancreatic fistulas (ISGPF) as the inability to return to a standard diet by the end of the first postoperative week and includes prolonged nasogastric intubation of the patient [19].

Both intra-abdominal drains were attached to vacuum containers and were monitored daily for volume output. Amylase levels from the serum and drainage fluid were measured on day-5. Any external fistula with a drain output of any measurable volume after postoperative day-3 with an amylase level greater than three times the upper limit of the normal serum value is defined as a postoperative PF according to the international study group for pancreatic fistulas (ISGPF) [20]. Bile leak according to the international study group of liver surgery was suspected if bilirubin concentration in the drain fluid at least three times the serum bilirubin concentration on or after postoperative day 3 or as the need for radiologic or operative intervention resulting from biliary collections or bile peritonitis (ISGLS) [21].

In total, 36 patients underwent PG including 21 males and 15 females. The mean age of this cohort was 58.44 years (SD 15.76). 62% of Patients had BMI = 25 (7 are missing). 30.6% of patients had been diagnosed with diabetes mellitus preoperatively and already started on treatment.

Thirty six percent of patients had biliary stents placed pre-operatively via endoscopic retrograde cholangiopancreatography (ERCP) or percutaneous drainage (PTC) of the biliary system under radiological guidance. Intraoperatively the pancreatic stump was soft and friable in 31.6% of the cases. Histopathological analysis confirmed 7 benign cases and 29 malignant cases as given in Table 1 . In this cohort, the mean tumor size was 3.29 cm (range, 0–11 cm)-malignant 3.57 cm in diameter vs. 2.66 cm for benign lesions, of the malignant cases vascular invasion was seen in 23.5% of specimens, for that patients received chemotherapy as per multidisciplinary meeting decision. Fibrosis presented in 41.7% of specimens. R0 resection achieved successfully in 52.8%, R1 resection in 19.4% and R2 resection in 5.6% of patients. The mean operating time was 5.6 hours (SD 1.74), the mean intraoperative blood loss was 676.39 ml (SD 1222.96). The mean hospital stay was 11.86 days (SD 6.05). The mean survival rate was 9.5 months (SD11.68).

During the follow-up period, four patients (11.1%) died 9 months to 4 years after the procedure. One patient was lost to follow up. In hospital mortality (define - is this 90 days) occurred in two patients (5.5%). One patient died from intraoperative bleeding as a result of a porta hepatis injury and the other died on day-1 due to DIC as a result of massive blood transfusion for intraoperative bleeding. Two patients were excluded from the final survival analysis.

The overall morbidity was 83.3% which included wound infections (22.2%), atelectasis (13.9%) and delayed gastric emptying and postoperative hemorrhage (8.3%). Abscesses and collections, wound dehiscence and PE occurred in a total of (5.6%) of the cases. Intraoperative hemorrhage, pancreatic and biliary fistulas, lymphatic duct injury and DVT account for 2.8% of the complications. The lowest rate was (0%) in major complications such as cholangitis and acute pancreatitis (Table 2). In this cohort, we noticed that there was no correlation between the pancreatic fistula and the development of intra-abdominal abscess or collection. Re-admission was needed in three patients, first patient readmitted after two weeks with recurrent vomiting, upper gastrointestinal scope was done for him and it was normal-no mechanical causes of obstruction. He improved with diet modification within few weeks. The second patient presented after three weeks from the discharge with hematemesis, upper gastrointestinal scope and CT angiogram were obtained for him and both of them were negative. Patient admitted for observation, his bleeding stopped spontaneously and patient discharged home. In the third case, patient readmitted within few days from the discharge with wound infection and sepsis, which needed frequent dressing and IV antibiotic. Steatorrhea was noticed in 13.9% of patients postoperatively in the first few visits which treated with recombinant pancreatic enzymes. On the long-term follow-up four of our patients died after 2–3 years from the operation and one of them lost his follow-up.

Periampullary tumors are defined as those arising from the head of the pancreas, the distal common bile duct and the duodenum. Overall they have a high mortality rate compared with other cancer types, and account for >30,000 cancer related deaths per year in the U.S. [22]. The most definite treatment for periampullary tumors is PD although this is still considered as a relatively risky procedure with a high morbidity and mortality rate. Post-operative morbidity occurs in up to 45.1% of cases [23]. The most significant complication is a pancreaticoenteric anastomotic leak which is reported to occur in 5–20% of cases [15]. The presence of a pancreatic leak increases the mortality rate, Lerut et al. reported that pancreatic fistula accounted for 55% of postoperative deaths [24]. PJ is the most commonly used technique for managing the pancreatic remnant. However, when using this technique anastomotic leaks and pancreatic fistulas have been reported in up to 11–40% of cases [25]. Pancreaticoenteric anastomotic failure often results in delayed gastric emptying, pancreatic fistula, and wound infection. The consequences of a pancreatic fistula may be dire sometimes, leading to intraabdominal hemorrhage, intraabdominal infection, wound dehiscence and even death [26]. Yeo et al. reported on outcomes after 650 consecutive PDs [27]. They noted that four percent of patients required repeat surgery and that one third of these patients died after the second operation as a result mostly related to leakage from the PJ anastomosis. The modified technique of PG used in the present series resulted in zero leaks and numerous other groups have similar good results using this technique, Miyagawa et al. [28] in a retrospective comparative study found that PG was superior to PJ in terms of anastomotic breakdown. Bartoli, [29] undertook a meta-analysis of pancreatic fistula and relative mortality in malignant disease after PD. They compared data on PG versus PJ anastomosis performed in three different ways, i.e., PJ end to side, PJ end to end, and Wirsung duct to jejunal end to side anastomosis. In all cases, PG was associated with lower morbidity and mortality rates compared with a PJ anastomosis. In a retrospective study on 214 patients Aranha GV et al. showed a significant difference in mortality between the two groups in favor of the PG group [27]. Anastomosies of the pancreas was first described theoretically by Tripodi and Sherwin in 1934 [30]. The technique was further evaluated in the laboratory by Person and Glenn [31]. The first successful PG in a patient was performed by Waugh and Clagett in 1946 [32]. In a review done from 1946 to 1997 by Mason, pancreaticogastrostomy sees to be safe procedure . A total of 199 cases using the PG technique were described in the literature. A mortality of 4.5% (9/199) and a leakage rate of 1% (2/199) were reported. None of the nine deaths described in this review were attributed to PG. During the time period 1991 to 1997, the number of PGs described in the literature was 614. The incidence of leakage for this period when compared with the previous period rose to 4.7% (29/614), with an overall mortality of 3.3% (20/614). In this mortality group only 15% of deaths were related to the PG anastomosis [33]. Numerous theories have been proposed as to why a PG might be more favorable in terms of the risk of an anastomotic leak compared with PJ. Firstly, the stomach is known to have a rich blood supply which might decrease the risk of ischemic complications. Secondly, the thick wall of the stomach makes the sutures hold better than in the jejunum. Thirdly, the pancreas lies immediately behind the stomach and anastomosing it in the posterior wall allows a tension free anastomosis which may reduce the risk of leakage. Fourthly, there is no accumulation of pancreaticobiliary secretions as presumably occurs in a long loop of the jejunum which can add tension to the anastomosis. Fifthly, the relatively acidic media of the stomach do not activate the pancreatic proteolytic enzymes which may also help protect the anastomosis. Lastly, the alkaline pancreatic secretions neutralize the gastric acidity which protects against potential marginal ulceration [23] [24] .

In addition to these, there are other controllable factors which can affect the anastomosis and the general outcome. These can be divided into: pancreatic factors, patient factors and operative factors. Pancreatic factors include [1] the pancreatic texture where the fibrotic pancreatic remnant is less likely to leak after an anastomosis compared with a soft and fragile pancreatic remnant; [2] the original pathology; [3] the blood supply to the pancreatic remnant; [4] and the pancreatic duct size. Patient related factors include: age, gender, level of preoperative jaundice, and comorbid diseases. Operative factors include operation time, blood loss, type of anastomosis and stenting of the pancreatic duct [34] [35].

Numerous pharmacological agents have been described to decrease or prevent pancreatic anastomotic leaks. The most commonly used medication is octreotide. A recent meta-analysis by Connor et al., showed that octreotide usage reduces the incidence of postoperative complications in pancreatic surgery [24].

Pancreatic stump invagination into the gastric lumen for about 2 cm with a two layer fixation of the ventral surface of the pancreas was associated with low risk of pancreatic anastomosis failure in our study. The limitations of this include the small sample size and the range of different pathologies in this series. Larger multi-centric studies are needed to establish whether pancreaticogastrostomy reconstruction following pancreaticoduodenectomy is superior to pancreaticojejunal reconstruction.

Special appreciation is offered to Dr. Thomas J. Hugh for reviewing and modifying the paper.

1. Melroy A D'souza, Shailesh V Shikhandi. Pancreatic resectional surgery: an evidence-based perspective. J Cancer Res Ther 2008 Apr-Jun;4(2):77–83.   [Pubmed]    
2. Wagner M, Redaelli C, Lietz M, Seiler CA, Friess H, Büchler MW. Curative resection is the single most important factor determining outcome in patients with pancreatic adenocarcinoma. Br J Surg 2004 May;91(5):586–94.   [CrossRef]   [Pubmed]    
3. Smeenk HG, Tran TC, Erdmann J, van Eijck CH, Jeekel J. Survival after surgical management of pancreatic adenocarcinoma: does curative and radical surgery truly exist? Langenbecks Arch Surg 2005 Apr;390(2):94–103.   [CrossRef]   [Pubmed]    
4. Shrikhande SV, Barreto G, Shukla PJ. Pancreatic fistula after pancreaticoduodenectomy: the impact of a standardized technique of pancreaticojejunostomy. Langenbecks Arch Surg 2008 Jan;393(1):87–91.   [CrossRef]   [Pubmed]    
5. Arnaud JP, Tuech JJ, Cervi C, Bergamaschi R. Pancreaticogastrostomy compared with pancreaticojejunostomy after pancreaticoduodenectomy. Eur J Surg 1999 Apr;165(4):357–62.   [Pubmed]    
6. Goldsmith HS, Ghosh BC, Huvos AG. Ligation versus implantation of the pancreatic duct after pancreaticoduodenectomy. Surg Gynecol Obstet 1971 Jan;132(1):87–92.   [Pubmed]    
7. Papachristou DN, Fortner JG. Pancreatic fistula complicating pancreatectomy for malignant disease. Br J Surg 1981 Apr;68(4):238–40.   [CrossRef]   [Pubmed]    
8. Di Carlo V, Chiesa R, Pontiroli AE, et al. Pancreatoduodenectomy with occlusion of the residual stump by Neoprene injection. World J Surg 1989 Jan-Feb;13(1):105–10; discussion 110–1.   [CrossRef]   [Pubmed]    
9. Suc B, Msika S, Fingerhut A, et al. Temporary fibrin glue occlusion of the main pancreatic duct in the prevention of intra-abdominal complications after pancreatic resection: prospective randomized trial. Ann Surg 2003 Jan;237(1):57–65.   [CrossRef]   [Pubmed]    
10. Strasberg SM, Drebin JA, Mokadam NA, et al. Prospective trial of a blood supply-based technique of pancreaticojejunostomy: effect on anastomotic failure in the Whipple procedure. J Am Coll Surg 2002 Jun;194(6):746–58; discussion 759–60.   [Pubmed]    
11. Lillemoe KD, Cameron JL, Kim MP, et al. Does fibrin glue sealant decrease the rate of pancreatic fistula after pancreaticoduodenectomy? Results of a prospective randomized trial. J Gastrointest Surg 2004 Nov;8(7):766–72; discussion 772–4.   [CrossRef]   [Pubmed]    
12. Warshaw AL, Thayer SP. Pancreaticoduodenectomy. J Gastrointest Surg 2004 Sep-Oct;8(6):733–41.   [CrossRef]   [Pubmed]    
13. Sakorafas GH, Friess H, Balsiger BM, Büchler MW, Sarr MG. Problems of reconstruction during pancreatoduodenectomy. Dig Surg 2001;18(5):363–9.   [CrossRef]   [Pubmed]    
14. Sutton CD, Garcea G, White SA, et al. Isolated Roux-loop pancreaticojejunostomy: a series of 61 patients with zero postoperative pancreaticoenteric leaks. J Gastrointest Surg 2004 Sep-Oct;8(6):701–5.   [CrossRef]   [Pubmed]    
15. Papadimitriou JD, Fotopoulos AC, Smyrniotis B, Prahalias AA, Kostopanagiotou G, Papadimitriou LJ. Subtotal pancreatoduodenectomy: use of a defunctionalized loop for pancreatic stump drainage. Arch Surg 1999 Feb;134(2):135–9.   [CrossRef]   [Pubmed]    
16. Ingebrigtsen R, Langfeldt E. Pancreaticogastrostomy. Lancet 1952 Aug 9;2(6728):270–1.   [CrossRef]   [Pubmed]    
17. Johnson CD. Pancreaticogastrostomy after resection of the pancreatic head. In: Malfertheier P, Buchler M, Beger HG eds. Standards in Pancreatic Surgery. Berlin: Springer-Verlag; 1993. p. 663–75.    
18. Bassi C, Falconi M, Molinari E, et al. Reconstruction by pancreaticojejunostomy versus pancreaticogastrostomy following pancreatectomy: results of a comparative study. Ann Surg 2005 Dec;242(6):767–71, discussion 771–3.   [CrossRef]   [Pubmed]    
19. Wente MN, Bassi C, Dervenis C, et al. Delayed gastric emptying (DGE) after pancreatic surgery: a suggested definition by the International Study Group of Pancreatic Surgery (ISGPS). Surgery 2007 Nov;142(5):761–8.   [CrossRef]   [Pubmed]    
20. Bassi C, Dervenis C, Butturini G, et al. Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery 2005 Jul;138(1):8–13.   [CrossRef]   [Pubmed]    
21. Koch M, Garden OJ, Padbury R, et al. Bile leakage after hepatobiliary and pancreatic surgery: a definition and grading of severity by the International Study Group of Liver Surgery. Surgery 2011 May;149(5):680–8.   [CrossRef]   [Pubmed]    
22. Fernandez-Cruz L. Periampullary carcinoma. In: Holzheimer RG, Mannick JA eds. Surgical Treatment: Evidence-Based and Problem-Oriented. Munich, Germany: Zuckschwerdt; 2001.    
23. Fang WL, Shyr YM, Su CH, Chen TH, Wu CW, Lui WY. Comparison between pancreaticojejunostomy and pancreaticogastrostomy after pancreaticoduodenectomy. J Formos Med Assoc 2007 Sep;106(9):717–27.   [CrossRef]   [Pubmed]    
24. Lerut JP, Gianello PR, Otte JB, Kestens PJ. Pancreaticoduodenal resection. Surgical experience and evaluation of risk factors in 103 patients. Ann Surg 1984 Apr;199(4):432–7.   [Pubmed]    
25. Trede M, Schwall G. The complications of pancreatectomy. Ann Surg 1988 Jan;207(1):39–47.   [CrossRef]   [Pubmed]    
26. Aranha GV, Hodul P, Golts E, Oh D, Pickleman J, Creech S. A comparison of pancreaticogastrostomy and pancreaticojejunostomy following pancreaticoduodenectomy. J Gastrointest Surg 2003 Jul-Aug;7(5):672–82.   [CrossRef]   [Pubmed]    
27. Yeo CJ, Cameron JL, Sohn TA, et al. Six hundred fifty consecutive pancreaticoduodenectomies in the 1990s: pathology, complications, and outcomes. Ann Surg 1997 Sep;226(3):248–57; discussion 257–60.   [CrossRef]   [Pubmed]    
28. Miyagawa S, Makuuchi M, Lygidakis NJ, et al. A retrospective comparative study of reconstructive methods following pancreaticoduodenectomy--pancreaticojejunostomy vs. pancreaticogastrostomy. Hepatogastroenterology 1992 Oct;39(5):381–4.   [Pubmed]    
29. Bartoli FG, Arnone GB, Ravera G, Bachi V. Pancreatic fistula and relative mortality in malignant disease after pancreaticoduodenectomy. Review and statistical meta-analysis regarding 15 years of literature. Anticancer Res 1991 Sep-Oct;11(5):1831–48.   [Pubmed]    
30. Tripodi AM, Sherwin CF. Experimental transplantation of the pancreas into the stomach. Arch Surg 1934;28(2):345–56.   [CrossRef]    
31. Person Jr EC, Glenn F. Pancreaticogastrostomy, experimental transplantation of the pancreas into the stomach. Arch Surg 1939;39:530–50.    
32. Waugh JM, Clagett OT. Resection of the duodenum and head of the pancreas for carcinoma; an analysis of thirty cases. Surgery 1946 Aug;20:224–32.   [Pubmed]    
33. Mason GR. Pancreatogastrostomy as reconstruction for pancreatoduodenectomy: review. World J Surg 1999 Mar;23(3):221–6.   [CrossRef]   [Pubmed]    
34. Cullen JJ, Sarr MG, Ilstrup DM. Pancreatic anastomotic leak after pancreaticoduodenectomy: incidence, significance, and management. Am J Surg 1994 Oct;168(4):295–8.   [CrossRef]   [Pubmed]    
35. Connor S, Alexakis N, Garden OJ, Leandros E, Bramis J, Wigmore SJ. Meta-analysis of the value of somatostatin and its analogues in reducing complications associated with pancreatic surgery. Br J Surg 2005 Sep;92(9):1059–67.   [CrossRef]   [Pubmed]