Salvage pancreaticogastrostomy for pancreatic fistulae ... - Springer Link

0 downloads 0 Views 106KB Size Report
Aug 28, 2012 - Indian Society of Gastroenterology 2012. Abstract Pancreatic anastomotic dehiscence after pan- creaticoduodenectomy (PD) remains a ...
Indian J Gastroenterol (September–October 2012) 31(5):263–266 DOI 10.1007/s12664-012-0213-1

CASE SERIES

Salvage pancreaticogastrostomy for pancreatic fistulae after pancreaticoduodenectomy Sanjay Govil

Received: 15 November 2011 / Accepted: 10 June 2012 / Published online: 28 August 2012 # Indian Society of Gastroenterology 2012

Abstract Pancreatic anastomotic dehiscence after pancreaticoduodenectomy (PD) remains a common problem. Although the management of this condition is mostly conservative, some patients require surgical intervention. This study reviews our experience with surgical intervention in this clinical setting. All patients who underwent PD by the author between 1999 and 2011 were reviewed. The causes for reoperation and mortality after PD were evaluated. The nature of the operative intervention and outcome in those who underwent reoperation for postoperative pancreatic fistula were analyzed. Reoperation was necessary in a total of 36/208 patients in this series and the overall mortality for the entire series was 6.25 % (13/ 208). Twelve of these 36 reoperations in 208 patients were for treatment of pancreatic anastomotic dehiscence after PD. Five (42 %) patients reoperated for anastomotic dehiscence died, including four of six patients that underwent surgical drainage of percutaneously inaccessible collections and one of two patients that underwent completion pancreatectomy. None of those who underwent salvage pancreaticogastrostomy (PG) died, nor did they require additional interventions prior to discharge from hospital. In our experience, salvage PG was an effective and organ function preserving technique to manage pancreatic anastomotic dehiscence after PD. Keywords Anastomotic leak . Pancreas . Surgery

S. Govil (*) Bangalore Institute of Oncology Specialty Centre, HCG Towers, #8 P. Kalinga Rao Road, Sampangiram Nagar, Bangalore 560 027, India e-mail: [email protected]

Introduction Despite significant reduction in mortality associated with pancreaticoduodenectomy (PD) over the last few decades, pancreatic fistula occurs in 2 % to 28 % of patients [1]. In 2005 postoperative pancreatic fistulae were graded into 3 grades (Table 1) by the International Study Group for Pancreatic Fistula (ISGPF) [2]. There is no consensus in the management of postoperative pancreatic fistula but most patients with this potentially fatal condition are managed without operation [1, 3–5]. Surgery is sometimes necessary for persistent or progressive sepsis, hemorrhage or non-resolution. Procedures undertaken at reoperation range from abscess drainage to dismantling of the anastomosis with pancreatic duct occlusion [6], bridge-stent anastomosis [7, 8], revision pancreaticoenterostomy [9] or completion pancreatectomy [6, 10]. In view of the infrequent need for reoperation most series consist of a small number of patients. This study evaluates the management of pancreatic fistulae requiring reoperation (Grade C) in a cohort of 208 pancreaticoduodenectomies performed by the author and discusses the risks and benefits of different surgical techniques.

Methods Medical reports of 208 pancreaticoduodenectomies performed by the author over a 12 year period between 1999 and 2011 were reviewed to identify all patients who underwent reexploration for management of pancreatic leaks. All patients underwent pancreaticojejunostomy (PJ) onto the same jejunal limb as the hepaticojejunostomy and gastroduodenojejunostomy. The pancreaticojejunal anastomosis

264 Table 1 Grading of postoperative pancreatic fistulae by the International Study Group on Pancreatic Fistula

a

Antibiotics, percutaneous drainage or octreotide

b

POPF is defined as any measureable drain fluid on or after day 3 with amylase content greater than 3 times normal POPF postoperative pancreatic fistulae

Indian J Gastroenterol (September–October 2012) 31(5):263–266

Clinical condition Specific treatmenta USG/CT if obtained Persistent drainage after 3 week Reoperation Death related to POPFb Signs of infection Sepsis Readmission

was performed with 2 layers of 4/0 Prolene and was the first anastomosis on the jejunum. The anastomosis was internally stented whenever the main pancreatic duct was less than 5 mm in diameter. Patients routinely underwent feeding jejunostomy and did not receive perioperative octreotide. Contrast enhanced CT scan was performed on all patients who had bile stained effluent from the abdominal drain, had tachycardia and leukocytosis 7 days after operation or had not tolerated normal diet by 7 days after operation. Abdominal collections were radiologically drained whenever possible. Surgical drainage was performed when radiological access to clinically significant intraabdominal collections was not possible. Only patients who were operated upon and in whom the diagnosis of pancreaticojejunal leak was confirmed at laparotomy were included in this study. The records of all patients who underwent radiological intervention or died after surgery were also evaluated. Particular note was made of the reason and timing of reexploration, the procedure performed during reexploration and the outcome. Anastomotic leaks were managed with antibiotics, therapeutic trial of octreotide, jejunal feeding and radiological drainage of collections. Completion pancreatectomy was performed if the CT showed pancreatic necrosis. After 2006 all patients who were reexplored for a pancreatic leak underwent PG. The PG was performed after dismantling the anastomosis and staple-transecting the end of the jejunum. The pancreas was then cut back to healthy tissue when necessary and mobilized off the splenic vein. A small gastrotomy, approximately half the diameter of the pancreas was made at an appropriate site on the posterior wall of the stomach and the mobilized pancreas was inserted through this into the stomach. A few interrupted sutures were taken between the pancreatic capsule and the seromuscular layer of the stomach. When necessary an anterior gastrotomy was made to place additional sutures. The pancreatic bed was drained and the stomach kept decompressed with a nasogastric tube postoperatively. Postoperative feeding was through a jejunostomy tube.

Grade A

Grade B

Grade C

Well None Negative No No No No No No

Often well Sometimes Negative/Positive Usually yes No No Yes No Possibly yes

Ill-looking Always Positive Yes Yes Possibly yes Yes Yes Possibly yes

Results Twelve of 208 patients needed exploration to manage pancreaticojejunal leaks after PD, Table 2. Surgical intervention was performed at a median of 14.5 days (range 12–33 days) after PD. Two patients underwent completion pancreatectomy because of extensive pancreatic necrosis. Prior to 2006, all six patients that were reexplored underwent external drainage of the leak with two survivors. Four of these patients needed further interventions in the form of abscess drainage or embolization for hemorrhage. After 2006, all four patients who were reexploration for pancreaticojejunal

Table 2 Nature, timing, indication and outcome of reoperation for International Study Group on Pancreatic Fistula Grade C fistulae after pancreaticoduodenectomy Day

Indication

Reintervention

Drainage 1 19 sepsis yesa 2 16 sepsis yesa 3 18 sepsis yesb 4 17 sepsis yesb 5 14 sepsis no 6 33 sepsis no Completion pancreatectomy 1 12 sepsis no 2 20 sepsis no Salvage pancreaticogastrostomy 1 2 3 4 a

16 18 14 16

sepsis bleedd bleed sepsis

no noc no no

Diabetes

Outcome

yes no

dead alive dead dead dead alive

yes -

alive dead

no no yes no

alive alive alive alive

Angiography and embolization

b

Percutaneous abscess drainage

c

Percutaneous abscess drainage on day 11

d

Angiography unable to localize bleeding (mesenteric vessel)

Indian J Gastroenterol (September–October 2012) 31(5):263–266

leaks underwent dismantling of the pancreaticojejunostomy and conversion to PG. One patient needed ligation of a bleeding mesenteric vessel and another suture ligation of the gastroduodenal artery stump at laparotomy. None of these patients needed reintervention and all four survived. Nine patients needed percutaneous drainage (4) or aspiration (5) of postoperative intraabdominal collections and three needed radiological embolization of gastroduodenal artery bleeds but did not require surgical exploration. Accurate data on drain fluid amylase in the entire series was not available, making retrospective determination of Grade A and B leaks unreliable. An additional 12 patients needed reexploration for reasons other than pancreaticojejunal leakage. These included 2 for duodenojejunal leaks, 1 for bile peritonitis after hepaticojejunal leak, 1 for pancreatic stump bleeding, 2 for intestinal obstruction, 2 for intraperitoneal bleeding, 1 for a bleeding anastomotic ulcer, 2 for gangrene of the jejunum and one for a leak after jejunostomy removal. Apart from the 5 deaths related to pancreaticojejunal leaks there were 8 additional deaths (gastroduodenal artery pseudoaneurysm bleed 1, anastomotic ulcer bleed 1, pancreatic stump bleed 1, unrecognized celiac artery obstruction 1, hemorrhage related to severe intraoperative blood loss due to portal vein injury 1, chylous ascites 1, intracranial hemorrhage 1 and aspiration pneumonia 1), bringing the total to 13 of 208 patients or 6.25 %.

Discussion Pancreaticojejunal anastomotic leaks after PD are associated with a mortality of 20 % to 88 % [6]. In the present series, the overall mortality was 6.25 %, with 38 % of the mortality related to pancreatic leaks. Forty-two percent of those who were reoperated to manage pancreatic leaks died. The correct way to manage pancreatic leaks has been controversial but the consensus appears to be to remain as conservative as possible [3, 11, 12]. Most anastomotic leaks are managed with antibiotics, drainage of collections to create controlled fistulae, and adequate nutrition. Octreotide has not provided any benefit in the management of these anastomotic leaks although it is often used [13]. The controlled pancreatic fistula is then managed with gradual withdrawal of the drain over a period of a few weeks [12]. However, some pancreatic leaks require additional intervention, most commonly due to hemorrhage, but also due to persistent sepsis from undrained collections or nonresolving high-output fistulae [14]. Surgical intervention in this clinical setting is difficult and has a poor outcome. Since this clinical situation is infrequent, most published series include small numbers of patients making it difficult to draw firm conclusions or establish guidelines. At best they can state that a particular

265

course of action is safe and effective. The present series is no different in this regard. Any surgery in this clinical setting can further disrupt the anastomosis. Simple drainage after surgery results in a deadly soup of biliary, pancreatic and intestinal secretions bathing the retroperitoneum, resulting in hemorrhage or worsening sepsis. Four of six of our patients treated in this fashion died with most needing further radiological interventions. van Berge Henegouwen et al. performed completion pancreatectomy for pancreatic fistulae at a median of 5 days after the PD (range 1–20 days) [6, 10] and found this operation to be associated with fewer reinterventions, shorter hospital stay and no mortality. Others, who re-intervened after a longer period of conservative management [15] or in the setting of generalized peritonitis [16], were unable to achieve the same degree of success and experienced operative mortality rates of 24 % to 62 %. Completion pancreatectomy also necessitated splenectomy, occasionally total gastrectomy, and inevitably resulted in brittle diabetes. One of the two patients in this series who underwent completion pancreatectomy died 3 days later from ongoing sepsis and multi-organ failure, while the other died 16 months later from hypoglycemia. Managing pancreatic insufficiency in India is a difficult and expensive proposition. Completion pancreatectomy should probably be limited to clinical situations in which there is no alternative, such as when there is extensive pancreatic necrosis due to postoperative pancreatitis. Other techniques of managing the remnant pancreas after dismantling the PJ are to ligate or occlude the main pancreatic duct [6], to create a bridge stent anastomosis between the pancreatic stump and jejunum [7, 8] or create a PG [9]. Main pancreatic duct ligation or occlusion is unreliable, may be associated with pancreatitis and hemorrhage, and has the highest mortality when compared to completion pancreatectomy or external drainage [6, 17]. The bridgestent anastomotic technique [7, 8] has been shown to be effective but relies on patency and adequacy of drainage by a stent bridging the distance between the pancreatic stump and the bowel. Salvage PG was reported by Bachellier et al. in 2008 [9]. They reported on 12 patients with PJ leaks after PD, 8 of whom were treated by completion pancreatectomy a median of 8 days postoperatively, and 4 treated by PG at a median of 10 days postoperatively. All 4 of those treated by PG survived compared to only 50 % of those who underwent CP. The need for further reinterventions and the incidence of postoperative diabetes was also less after PG although there was no significant difference in postoperative stay. Salvage PG has the benefit of being technically easy because of the location of the pancreas and stomach. Concerns about creating a fresh gastrotomy in the setting of infection are unfounded because of the thickness and vascularity of the stomach wall. The procedure is obviously easier after a

266

pylorus preserving PD but is technically feasible even after classical Whipple operation. Mobilization of the pancreas from the splenic vein is required but is not difficult as venous tributaries are often thrombosed. Creation of a small gastrotomy helps to hold the invaginated pancreatic stump within the stomach even without the need for sutures, although some may be placed if possible. Nasogastric decompression of the stomach postoperatively permits good drainage of pancreatic juice, which remains inactive due to the acidity of the gastric milieu. In the present series, salvage PG resulted in excellent resolution of pancreatic fistulae with lower mortality and fewer reinterventions than drainage procedures. Completion pancreatectomy was performed only in the setting of pancreatic necrosis when PG would not have been feasible but was associated with mortality and pancreatic insufficiency. With the increasing role of conservatism and radiological intervention in the management of pancreatic leaks after PD, the timing of surgical intervention is likely to be delayed. Completion pancreatectomy has been shown to be a hazardous operation in this setting [15]. Salvage PG may be the preferred operation when reoperation is necessary to treat pancreaticojejunal leaks after PD.

References 1. Haddad LBP, Scatton O, Randone B, et al. Pancreatic fistula after pancreaticoduodenectomy: the conservative treatment of choice. HPB. 2009;11:203–9. 2. Bassi C, Dervenis C, Butturini G, et al. Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery. 2005;138:8–13. 3. Kazanjian KK, Hines OJ, Eibl G, Reber HA. Management of pancreatic fistulas after pancreaticoduodenectomy: results in 437 patients. Arch Surg. 2005;140:849–54. 4. Veillette G, Dominguez I, Ferrone C, et al. Implications and management of pancreatic fistulas after pancreaticoduodenectomy. Arch Surg. 2008;143:476–81.

Indian J Gastroenterol (September–October 2012) 31(5):263–266 5. Fuks D, Piessen G, Huet E, et al. Life threatening postoperative pancreatic fistula (Grade C) after pancreaticoduodenectomy: incidence, prognosis and risk factors. Am J Surg. 2009;197:702–9. 6. De Castro SMM, Busch ORC, van Gulik TM, Obertop H, Gouma DJ. Incidence and management of pancreatic leakage after pancreaticoduodenectomy. Br J Surg. 2005;92:1117–23. 7. Kent TS, Callery MP, Vollmer CM Jr. The bridge stent technique for salvage of pancreaticojejunal anastomotic dehiscence. HPB. 2010;12:577–82. 8. Xu J, Dai X, Bu X, Gao F, Zhang X. Pancreaticojejunal bridge anastomosis: a novel option for surgeon to preserve pancreatic body and tail in urgent reoperation for intra-abdominal massive haemorrhage after pancreaticoduodenectomy. World J Surg. 2010;34:2457–62. 9. Bachellier P, Oussoultzoglou E, Rosso E, et al. Pancreaticogastrostomy as a salvage procedure to treat severepostoperative pancreatic fistula after pancreaticoduodenectomy. Arch Surg. 2008;143:966–70. 10. Van Berge Henegouwen MI, De Wit LT, van Gulik TM, Obertop H, Gouma DJ. Incidence, risk factors and treatment of pancreatic leakage after pancreaticoduodenectomy: drainage versus resection of the pancreatic remnant. J Am Coll Surg. 1997;185:18–24. 11. Munoz-Bongrand N, Sauvanet A, Denys A, Sibert A, Vilgrain V, Belghiti J. Conservative management of pancreatic fistula after pancreaticoduodenectomy with pancreaticogastrostomy. J Am Coll Surg. 2004;199:198–203. 12. Veillette G, Dominguez I, Ferrone C, et al. Implications and management of pancreatic fistulas following pancreaticoduodenectomy: the Massachusetts General Hospital experience. Arch Surg. 2008;143:476–81. 13. Cullen JJ, Sarr MG, Ilstrup DM. Pancreatic anastomotic leak after pancreaticoduodenectomy: incidence, significance and management. Am J Surg. 1994;168:295–8. 14. Shrikhande S, D’Souza M. Pancreatic fistula after pancreatectomy: evolving definitions, preventive strategies and modern management. World J Gastroenterol. 2008;14:5789–96. 15. Farley DR, Schwall G, Trede M. Completion pancreatectomy for surgical complications after pancreaticoduodenectomy. Br J Surg. 1996;83:176–9. 16. Gueroult S, Parc Y, Duron F, Paye F, Parc R. Completion pancreatectomy for postoperative peritonitis after pancreaticoduodenectomy. Early and late outcome. Arch Surg. 2004;139:16–9. 17. Benzoni E, Zompicchiatti A, Saccamano E, et al. Postoperative complication linked to pancreaticoduodenectomy. An analysis of pancreatic stump management. J Gastrointestin Liver Dis. 2008;17:43–7.