Laparoscopic surgery in pancreatic diseases: Pushing the boundaries

Abstract

Background

Laparoscopic surgery has expanded exponentially in the last two decades but, somehow it is limited in pancreatic surgery by virtue of the pancreas being a friable, retroperitoneal organ with difficult access and adjacent major vessels risking torrential bleed. It is thought to be unforgiving if not handled well. However, improvements in technology and surgeon's expertise have pushed the boundaries of minimal access surgery (MAS) to include pancreas in its domain. We present our series of laparoscopic pancreatic surgery (LPS) with an aim to look at the feasibility and outcomes.

Methods

This is a retrospective review of all LPS done at the Gastrointestinal Surgery (GIS) centre of a tertiary care Armed Forces Hospital over a period of 3 years.

Results

A total of 24 LPS were done during this period. The median age of the patients was 46 years (range; 13–81). There were 14 male and 10 female patients. Nine patients had at least one co-morbidity. Three patients underwent laparoscopic lateral pancreaticojejunostomy, 4 distal pancreatectomy, 4 laparoscopic Whipples pancreaticoduodenectomy, 6 laparoscopic pancreatic necrosectomy, 6 laparoscopic cystogastrostomy and 1 roux en y cystojejunostomy.

Conclusion

LPS can be performed for almost all open pancreatic surgeries and can be done with reasonable outcomes. However, it has a steep learning curve and therefore, a hybrid approach leading to a totally laparoscopic approach may be the way forward.

Introduction

Laparoscopic surgery has expanded exponentially in the last two decades but, somehow it is limited in pancreatic surgery by virtue of the pancreas being a friable, retroperitoneal organ with difficult access and adjacent major vessels risking torrential bleed. It is thought to be unforgiving if not handled well. However, improvements in technology and surgeon's expertise have pushed the boundaries of minimal access surgery (MAS) to include pancreas in its domain.

The spectrum of laparoscopic pancreatic surgery (LPS) ranges from benign to malignant and inflammatory to infective conditions. LPS have been found to be safe and advantageous in the experienced hands. We present our series of LPS, done at a tertiary centre with an aim to look at the feasibility and its outcomes and briefly discuss the literature on LPS.

Material and methods

This was a retrospective review of all patients who underwent LPS at a tertiary care Armed Forces Hospital between Jan 2014 and Apr 2017. The patient records and operative videos were perused. The demographic and clinical data, intra-operative factors and post-operative data were recorded. A minimum of 6 months’ follow-up was done.

The brief operative steps of the various procedures are described.

Laparoscopic pancreatic necrosectomy (LPN)

Patient was placed in lithotomy position and in a reverse Trendelenburg position, with the surgeon standing between the legs. Ports were placed keeping in mind, the location of the necrosis. However, in most cases the camera port was placed just above/through the umbilicus using optical, blunt trocar. Two 10 mm working ports were placed in mid-clavicular line (MCL), 5 cm above the umbilicus on either side. A right subcostal 5 mm port was placed in anterior axillary line (AAL) for retraction. Additional 5 mm ports were added as per the requirement. Necrotic pancreatic collections were confirmed by needle aspiration and along the needle, the cavity was opened using either harmonic scalpel or cautery hook. Immediately the suction was introduced into the small opening and liquid component aspirated to prevent contamination. Thereafter, the opening was widened. Using stone holding laparoscopic forceps, the pancreatic necrosum, which is loose and fluffy was taken out. Care was taken not to forcibly pull any necrosis or strand, as it could have led to torrential bleeding. The cavity was closed over two wide drains after a thorough wash. One drain was placed in the pelvis. Drains were preferably brought out using the port sites. The drains in the cavity were used for continuous saline lavage.

Laparoscopic lateral pancreaticojejunostomy (LLPJ)

Positioning of the patient and ports placement was almost the same as described in pancreatic necrosectomy except that, the left MCL port was 5 mm and right MCL port 12 mm for stapler firing. A Natheson retractor was used for liver retraction through the epigastric port (Fig. 1D). Patients with only MPD stones and MPD diameter of around 10 mm or more were chosen for laparoscopic surgery (Fig. 1A). After opening the lesser sac, the MPD was identified by noting a thinned out, compressible bulge in the anterior part of atrophic pancreas, which was confirmed by aspirating clear fluid. Along the needle, using hook cautery, the MPD was opened in its entire length, falling just 0.5 cm short of duodenum and tip of the tail of pancreas and all calculi cleared (Fig. 1B). Laparoscopic stapler was used to divide the jejunum, 20 cm from the DJ flexure. Stapled jejuno-jejunostomy was done keeping the roux limb of around 50 cm. The roux limb was brought up by retrocolic route. Jejunum was opened along the anti-mesenteric border and anastomosed with the opened MPD using 2/0 PDS continuous in single layer (Fig. 1C). Drain was placed next to LPJ.

Fig. 1.

(A) Axial CT scan showing dilated MPD filled with stones. (B) Intra-operative photograph showing extraction of stone from main pancreatic duct (MPD). (C) Intra-operative photograph showing lateral pancreaticojejunostomy in progress. (D) Post-operative photograph showing ports & drain position.

Laparoscopic pancreatic cystogastrostomy (LPG)

Positioning of the patient and ports placement was exactly the same as in LLPJ. Anterior gastrostomy of 5–6 cm, was made using harmonic/hook cautery at the most prominent part of the bulge. Needle aspiration was done through the posterior stomach wall to confirm the location of the cyst and then along the needle, using hook cautery, the cyst was entered and the pancreatic brownish black fluid was simultaneously aspirated without entering the cyst. The cystogastrostomy was made wide enough to allow the staple jaw to enter. A laparoscopic stapler 60 mm was used to make a long cystogastrostomy. Anterior gastrotomy was closed either by stapler or 2/0 PDS suture.

Laparoscopic distal pancreatectomy (LDP)

The patient was positioned in the right lateral position with 40° tilt and the left arm abducted (Fig. 2B). The telescope was inserted through the subcostal port in AAL using 12 mm optical port. Pneumoperitoneum was created. The camera port 10 mm through umbilicus and another 10 mm port in the epigastrium for retraction were placed. A left 5 mm working port at MCL between the camera and the epigastrium port (Fig. 2D) was also placed. In case of pancreaticosplenectomy and Warshaw technique, after opening the lesser sac, the splenic artery and vein were divided followed by pancreatic mobilization (Fig. 2C). However, in case of spleen preserving distal pancreatectomy, the splenic vessels were spared. Specimen was removed in a bag after extending the incision of the 12 mm port to 5 cm.

Fig. 2.

(A) CECT axial scan showing a hypodense focal lesion with internal septations in the body of pancreas resulting in focal contour bulge. (B) Partial thoracoabdominal position with left side up. (C) Intraoperative picture of laparoscopic distal pancreaticosplenectomy. (D) Post-operative picture showing ports & drain position.

Laparoscopic Whipples pancreaticoduodenectomy (LPD)

Patient was positioned in modified lithotomy position with reverse Trendelenburg position. A 10 mm camera port through umbilicus, 10 mm retraction port in epigastrium, 10 mm working port in right subcostal region in MCL and 12 mm port at left lumbar region in AAL were used. Two 5 mm working port in left AAL and MAL, 5 cm above the umbilicus (Fig. 3D) were added. Same steps were followed as in open Whipples pancreaticoduodenectomy (Fig. 3A). However, PG was done in 4 layers, invaginating the pancreatic stump in the stomach (Fig. 3B). In all patients, HJ was done with continuous 3/0 PDS as the ducts were dilated. Only in the second patient, laparoscopic feeding gastrostomy was done and pfannenstiel incision was taken to extract the specimen. In third and fourth patient, specimen was removed from 5 cm incision midway between umbilicus and epigastric port, and through the same incision feeding jejunostomy was done.

Fig. 3.

(A) Intra operative photograph showing uncinate dissection. (B) Intra operative photograph showing two layer gastrojejunostomy in progress. (C) Intra operative photograph showing two layer duct to mucosa pancreaticojejunostomy in progress. (D) Post-operative photograph showing ports & drain position.

Results

24 patients underwent LPS during this period. These included 6 pancreatic necrosectomies, 7 cysto-enterostomies, 3 pancreaticojejunostomy, 4 distal pancreatectomy and 4 Whipple's pancreaticoduodenectomies. The median age of the patients was 46 years (range; 13–81). There were 14 male and 10 female patients. Nine patients had at least one co-morbidity. Fifteen, 8 and 1 patient had an ASA grading of II, III and IV, respectively.

Laparoscopic pancreatic necrosectomy

We performed 6 laparoscopic pancreatic necrosectomy. All had severe acute pancreatitis with CT severity score of 10/10. They were initially managed by the gastroenterologist and had undergone multiple pigtail drainage with frequent upgradation. After reasonable time (more than 4 weeks), when the patient condition did not improve, the case was referred for surgery (Table 1). According to the location of the collection, different routes were chosen laparoscopically for necrosectomy. Optical port was used to enter the abdomen followed by extensive and meticulous adhesiolysis. The median blood loss and operative time was 100 ml and 150 min respectively. We had one conversion, the last case. This was due to duodenal necrosis detected during necrosectomy which was managed by repair over a T-tube and pyloric exclusion.

Table 1.

Clinical and outcome variables of patients with IPN.

	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6
Age/sex	76/F	46/M	30/M	37/M	52/M	32/M
Etiology	Idiopathic	Azathioprine	Alcohol + gall stones	Gall stones	Alcohol	Alcohol
Organ failure	None	01 (RS)	01 (RS)	03 (CVS, RS, ES)	03 (CVS, RS, ES)	None
Surgery from onset (weeks)	5th	8th	11th	8th	7th	7th
Approach	Transgastric	Transgastric	Paracolic	Infragastric	Paracolic	Infracolic
Blood loss (ml)	50	50	100	150	50	200
Operative time (min)	120	100	120	200	150	220
Post op events	Uneventful	Uneventful	Pancreatic fistula	Pancreatic fistula	Recovering well, later aspiration pneumonia and death	Duodenal leak, SSI and diabetes
Post operative Hospital stay (days)	6	10	74	92	87	50

CVS, cardiovascular system; RS, respiratory system; ES, excretory system.

Laparoscopic lateral pancreaticojejunostomy

Three patients underwent LLPJ for chronic pancreatitis. All three were female patients with stones in the main pancreatic duct with dilatation (Fig. 1A). We did not encounter any post-op complications and all patients are pain free at a median follow up of 18 months (range; 13–24) (Table 2).

Table 2.

Clinical and outcome variables of patients with chronic calcific pancreatitis.

	Patient 1	Patient 2	Patient 3
Age (years)	25	43	52
Sex	Female	Female	Female
Symptoms	Intractable pain	Intractable pain	Intractable pain, DM
Symptom duration (years)	7	9	12
MPD diameter (mm)	9	11	10
Blood loss (ml)	Minimal (<50)	Minimal	Minimal
Operative time (min)	150	180	190
Complications	Nil	Nil	Nil
Hospital stay (days)	4	5	4
Pain relief (%)	100	100	100
Follow up period (months)	24	18	13
Resolution of diabetes	–	–	Reduced insulin
Weight gain (kg)	7	5	4
Recurrence of pain	No	No	No

Laparoscopic pancreatic cystoenterostomy

We operated on 07 patients; all were symptomatic with vague upper abdominal pain and lump. One patient also had obstructive jaundice due to compression of cyst on to the CBD. The median operative time was 1 h and blood loss was 50 ml in patients who underwent Lap cystogastrostomy. Only one patient had complication in the form of upper GI Bleed requiring blood transfusion and endotherapy in the form of adrenaline injection to the bleeding cystogastrostomy margin. The patient, who underwent lap roux en y cystojejunostomy, had a large cyst projecting through the transverse mesocolon with minimal contact with stomach.

The median size of cyst was 15.6 cm. All patients are asymptomatic after a median follow up of 18.4 months (Table 3).

Table 3.

Clinical, operative and outcome variables of patients who underwent Laparoscopic cysto-enterostomy.

	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6	Case 7
Age	44	32	40	23	41	15	60
Sex	M	M	M	M	M	M	M
Symptom	Pain	Pain	Pain	Pain	Pain	Pain & vomiting	Pain, vomiting & Jaundice
Sign	Abd lump	Abd lump	Abd lump	Abd lump	Abd lump	Abd lump	Abd lump & icterus
Max diameter of cyst (cm)	10	17	13	12	23	15	19
Surgery	Cysto-gastrostomy (CG)	CG	CG	CG	CG	CG	Roux en Y cystojeju-nostomy
Blood loss (ml)	50	40	30	50	100	50	100
Operative time (min)	70	60	65	70	55	60	180
Complications	Nil	Nil	Nil	Nil	Bleed	Nil	Nil
Follow up period (months)	36	32	20	18	12	7	6

Laparoscopic distal pancreatectomy

At our center, we have done four laparoscopic distal pancreatectomies (Fig. 2A–D). The details of which are mentioned in Table 4. In these patients, tumor was either in the body or tail of pancreas. The relation of the tumor with the SMA was seen preoperatively on CT console which helped in deciding the line of transection on pancreas. In all four patients, cut margins were free of tumor. We were able to salvage spleen in two patients as the tumor was in the body. None of our patients had any significant morbidity and were discharged home in less than a week.

Table 4.

Clinical, operative and outcome variables of patients with tumors in pancreatic body/tail.

	Patient 1	Patient 2	Patient 3	Patient 4
Age (years)	45	55	13	69
Sex	F	F	F	M
Presentation	Incidental detection	Pain abdomen	Pain abdomen & vomiting	Repeated episodes of pancreatitis
Tumor size (cm)/type	4/SCN	5/MCN	6/SPN	3/IPMN
Spleen preservation	No	No	Yes (Kimura technique)	Yes (Warshaw technique)
Operative time (min)	200	150	180	170
Blood loss (ml)	100	150	100	150
Hospital stay (days)	4	4	3	6
Morbidity	Nil	Nil	Nil	Upper pole splenic infarction; transient DM
Pancreatic fistula	No	No	No	No
Follow up	30 months	24 months	8 months	6 months

SCN, serous cystadenoma; MCN, mucinous cystadenoma; SPN, solid pseudopapillary neoplasm; IPMN, intraductal papillary mucinous neoplasm; DM, diabetes mellitus.

Laparoscopic Whipples pancreaticoduodenectomy

We did four laparoscopic Whipples pancreaticoduodenectomy during this period. The details of the cases are given in Table 5.

Table 5.

Clinical, operative and outcome variables of patients with periampullary carcinoma.

	Patient 1	Patient 2	Patient 3	Patient 4
Age	81	43	39	49
Sex	M	F	M	F
Presentation	Jaundice	Pancreatitis & Jaundice	Jaundice	Jaundice
Tumor size	2 cm periampullary	3 cm periampullary	5 cm, periampullary	1.2 cm, periampullary
Preoperative Biliary Drainage	SEMS	Plastic stent	Direct surgery	Direct surgery
Surgery	Lap resection → open reconstruction	Lap WPD (PG)	Lap WPD (PJ)	Lap WPD (PG)
Operative time	430 min	480 min	550 min	540 min
Blood loss	500 ml	150 ml	200 ml	800
Length of stay	30 days	24 days	10 days	7 days
NG tube removal	POD 3	POD 4	POD 3	POD 2
Morbidity	Bulbar palsy	GJ leak	Nil	Nil
Pancreatic fistula	No	No	No	No
Histology	Adenocarcinoma Periampullary region LN – 12 harvested, margins negative	Adenocarcinoma ampulla (T2N0Mx) LN-15, margins negative	Adenocarcinoma ampulla (T2N1Mx) LN-2/18 positive, margins negative	Adenocarcinoma ampulla (T2N0Mx) LN-0/26positive, margins negative
Follow up	18 months	15 months	10 months	06 months

WPD, Whipples pancreaticoduodenectomy; PG, pancreato-gastrostomy; PJ, pancreato-jejunostomy.

All the patients whom we operated were diagnosed with periampullary Cancer. Our first patient underwent a self expanding metallic stent (SEMS) placement for pre-operative biliary drainage as he was not willing for surgery initially. We did the resection part of PD by laparoscopic surgery and did an open reconstruction through right subcostal incision. His stay in the hospital was prolonged because of bulbar palsy but he did not have any complications related to LPD. Our second patient had presented with biliary pancreatitis and Jaundice because of CBD stones. ERCP stenting followed by Lap cholecystectomy was done. While attempting to remove the stent, the gastroenterologist noted a growth at ampulla and took the biopsy. Stent was not removed. Following biopsy confirmation of adenocarcinoma, patient underwent TLPD. The operative time was increased by around 1 h due to dense adhesions as the result of acute pancreatitis. Pancreatogastrostomy (PG) was done in 4 layers (Fig. 3B). On the 5th day bile leak was noted from the epigastric port, though patient was stable. Laparoscopic Graham's repair of the leak from GJ was done using the previous port. She recovered well and was later discharged. Our third patient underwent upfront surgery without undergoing any drainage procedure. Large amount of visceral fat added to the operating time. Duct to mucosa pancreato-jejunostomy (PJ) was done (Fig. 3C). He had an uneventful recovery. Our fourth patient presented with obstructive jaundice and was taken for upfront surgery. However, in this patient blood loss was more, probably due to mild coagulopathy. None of our patient had delayed gastric emptying or pancreatic fistula. None required blood transfusion. Median lymph node harvest was 18 and margins were negative in all patients. All our patients are doing well.

Discussion

MAS has rapidly expanded and has touched all aspects of gastrointestinal surgeries (GIS). Having said this, LPS and laparoscopic liver resections (LLR) are considered as the final frontiers to be conquered in GIS. In fact, many consider a TLPD as a far more formidable surgery when compared to LLR. We, in our department, have embraced MAS surgeries and have made a steady progress to expand MAS to all subsets of GIS. Our experience with laparoscopic radical gastrectomies, advanced laparoscopic colorectal surgeries and laparoscopic radical cholecystectomies/choledochal cyst surgeries helped us to graduate from LPN to LDP and finally to LPD. We have done these surgeries with reasonably good outcomes.

Current status of LAPS

Many studies have shown that it is feasible to do almost all open surgery laparoscopically, but the question remains that though it is feasible, is it desirable? LPS is not accepted uniformly by HPB surgeons. However, it is slowly gaining acceptance by virtue of a few distinct advantages of LPS over open pancreatic surgery (OPS). Pancreatic surgery involves a large upper abdominal incision which has been shown to increase pulmonary complications. Since LPS is a MAS, the trauma caused to tissues in gaining access to the retroperitoneal organ is less which in turn causes less pain, promotes early mobilization and faster recovery of function and translates into a faster recovery and a shorter hospital stay.

Long term complication like incisional hernia and adhesive bowel obstruction are low. During surgery, though the operative time may be high but blood loss is less due to better magnification and therefore the need of blood transfusion is decreased.

Oncologically also it may add some benefit as the immunosuppressive effect is minimized due to the blunting of stress response¹ and less blood transfusion.²

Laparoscopic pancreatic necrosectomy

Recent literature shows that the morbidity and mortality is significantly lesser if step up approach is used in patients with infected pancreatic necrosis (IPN). We operated 6 consecutive patients with IPN by a laparoscopic approach. Though there are many MAS routes through which IPN can be tackled, we chose laparoscopic approach as we were most familiar and comfortable with this approach.

We were able to salvage five out of six patients of severe necrotizing pancreatitis with sepsis. The two most important factors which were responsible for this favorable outcome were – the timing of surgery and the minimal access route to pancreatic necrosectomy.

Currently the consensus is that surgery for infected pancreatic necrosis be delayed as long as possible, preferably to at least 3–4 weeks after onset of disease, to allow liquefaction and encapsulation of necrotic collection (walled of pancreatic necrosis).^{3, 4}

It has been shown in a landmark paper that the open approach to pancreatic necrosectomy was associated with high morbidity (34–95%) and mortality (11–39%) with a risk of long term pancreatic insufficiency whereas MAS reduces surgical trauma (i.e., tissue damage and systemic proinflammatory response) in an already critically ill patient leading to reduction in the incidence of new onset organ failure.^{5, 6}

Amongst other MAS for pancreatic necrosis, Laparoscopic pancreatic necrosectomy is not widely accepted as it carries a theoretical risk of dissemination of infection from the retroperitoneum to peritoneal cavity and possibility of enterotomy during adhesiolysis. However, the laparoscopic route has many advantages. In laparoscopy there is access to the entire abdomen which can address collections at inaccessible areas like paracolic gutter, root of mesentery etc., a cholecystectomy can be done simultaneously and laparoscopy does not invoke severe local or systemic response. A feeding jejunotomy can also be done. There is also reduction in the incidence of wound dehiscence and enterocutaneous fistula.⁷

Overall, laparoscopic necrosectomy has a clinical success rate of 70–95%, morbidity of approximately 20% and mortality of 0–18%.⁴

Laparoscopic lateral pancreaticojejunostomy

Chronic pancreatitis involves progressive fibrosis of the gland with loss of endocrine and exocrine function. Pain is the most common symptom, which if severe and intractable, requires surgery when non-surgical methods fail.

Surgery for chronic pancreatitis has evolved over time. In 1963, Partington et al. modified the Puestow-Gillesby pancreaticojejunostomy by creating an anastomosis between a longitudinally incised anterior surface of the pancreas and duct with a longitudinally incised Roux-en-Y jejunal loop.⁸ This is a decompressive surgery and is recommended in patients with refractory pain and an obstructed, dilated main pancreatic duct with no inflammatory mass or calcifications in the head of the pancreas.⁹ More than 90% patients are pain free after surgery.¹⁰ Laparoscopic LPJ has been found to be safe, effective and technically feasible procedure with all the advantages of MAS.¹¹

Laparoscopic cystoenterostomy

Pseudocyst pancreas is collection of fluid rich in pancreatic enzyme in the lesser sac secondary to the inflammation in pancreas. Its wall is made of granulation tissue. It is required to be drained if it is symptomatic with either pain and/or pressure symptoms. The options to drain a pancreatic Pseudocyst can be open or minimally invasive (either laparoscopic or endoscopic). And evidence today points to an increased advantage in employing the minimally invasive approach to this.¹² Laparoscopic cystoenterostomy is a simple, safe and fast with all the advantages of MAS.¹³

Laparoscopic distal pancreatectomy

Gagner et al. reported the first laparoscopic distal pancreatectomy (LDP) in 1996.¹⁴ Since then gradually laparoscopic distal pancreatectomy is gaining global acceptance.

Laparoscopic distal pancreatectomy is a feasible and safe alternative to open surgery in the treatment of benign and malignant tumor of pancreas, providing advantages in terms of reduced blood loss and enhanced postoperative recovery and the pancreatic fistula rate is no more than open surgery.¹⁵ Most comparative studies concluded that LDP results in fewer post operative complications and increases the chances of spleen preservation.^{16, 17} When oncological efficiency was compared, open and LDP were found to be comparable.¹⁶

Laparoscopic Whipples pancreaticoduodenectomy

Total laparoscopic pancreaticoduodenectomy (TLPD) has proven to be among the most advanced laparoscopic procedures. Few surgeons have performed this demanding procedure. This has largely been attributed to the time-consuming complexity of the operation and technical difficulties coupled with the need to develop advanced laparoscopic skills.

Relatively larger studies that have compared open to TLPD suggest that TLPD seems to be advantageous over open approach in terms of blood loss, rate of delayed gastric emptying, and hospital stay, but results in longer operative time.^{18, 19, 20, 21} This longer operative time is largely due to the reconstruction phase which is technically very demanding. Hence some authors argue in favor of a laparoscopic resection followed by an open reconstruction via a mini-laparotomy. Studies have looked into the cost effectiveness of TLPD and concluded that taking all factors including the operative time and hospital stay, both approaches were comparable when total costs were calculated. Further, TLPD provides better quality of life within the first 6 months after surgery.^{15, 22} Studies on laparoscopic versus open Whipple surgery specifically for cancer showed comparable R0 resection rates and lymph node retrieval.¹⁶ However, the fact remains that it has a steep learning curve. The way forward would probably be to adopt a hybrid approach and then graduate with experience to a TLPD.

Conclusion

In the last decade, due to the technological advancement and surgeons’ experience at centers of excellence for pancreatic surgery, LPS can be performed for almost all open pancreatic surgeries. LPS has shown all the advantages of minimally invasive surgery albeit, at the cost of time. It has proven benefit in pancreatic necrosectomy over open surgery as a part of step up approach. In oncological patients, LAPS has reached equal long term results compared to open surgery. We, at our centre, have been able to do LPS with reasonable outcomes. However further studies are required to demonstrate a convincing advantage. With the ever-improving technology, we can expect minimally invasive surgery to largely replace open pancreatic surgery in the coming time.

Conflicts of interest

The authors have none to declare.