Laparoscopic Hand-Assisted Parenchymal-Sparing Resections for Presumed Side-Branch Intraductal Papillary Mucinous Neoplasms

Introduction

The reported prevalence of malignancy in side-branch intraductal papillary mucinous neoplasm (IPMN) ranges from 6% to 46%, with a mean of 25%.^1,2 Additionally, the reported recurrence rate is up to 10% in patients undergoing resection with negative margins.^1,3 Established risk factors for malignancy include size >3 cm and the presence of a mural nodule.^4–6 These risk factors, in addition to symptoms such as obstructive jaundice or pain, are the usual indications for consideration of resection in patients with side-branch IPMN. Although significant improvements have been made, pancreatic surgery is still associated with significant morbidity and mortality. Clearly, the option of performing a less invasive resection with equivalent outcomes would be ideal.

With a lower risk of malignancy, we postulate that less invasive resections such as laparoscopic hand-assisted duodenal-sparing pancreatic head resection or enucleation can be safely performed in patients with side-branch IPMN. We describe our experience with 17 patients who underwent laparoscopic duodenal-sparing pancreatic head resection or enucleation with a preoperative diagnosis of side-branch IPMN at our institution.

Materials and Methods

The electronic medical record was queried for all patients admitted to our academic medical center who underwent pancreatic resection for a diagnosis of presumed IPMN between January 1, 2008 and March 30, 2013. Patients were selected who underwent laparoscopic hand-assisted pancreatic enucleation or duodenal-sparing pancreatic head resection. The medical records were reviewed, and data were extracted. Data collected included age, gender, presenting symptoms, preoperative imaging, and indications for surgery. Operative data collected included procedure performed, location of lesion, the use of intraoperative ultrasound, the treatment of the enucleated pancreatic surface, placement of drains, operative time, and estimated blood loss. Outcomes collected included pathology, 30-day complications, and length of hospital stay. Mean and median values were calculated. Complications were graded according to the Clavien–Dindo grading system, and pancreatic leak/fistula and delayed gastric emptying complications were also classified using the International Study Group of Pancreatic Surgery guidelines.^7–9 Postoperative imaging study results and dates were also collected.

Duodenal-sparing pancreatic head resections performed prior to 2010 underwent pancreaticogastrostomy reconstruction in two layers, whereas later resections underwent Roux-en-Y pancreaticojejunostomy in one or two layers. A single patient underwent a Roux-en-Y hepaticojejunostomy and pancreaticojejunostomy. With regard to drain management, the Jackson–Pratt drain fluid was checked for amylase level 24–48 hours after the surgery and removed if the value was less than three times the serum amylase level. If the Jackson–Pratt drain fluid amylase level was elevated, the drains were continued and removed when there was persistently low output (<30 mL/day) with good clinical condition of the patient.

Results

Seventeen patients were included in our study. The mean age of patients was 64 years old. Twelve patients were female, and 5 were male. The most common presenting symptom was abdominal pain (Table 1). All patients underwent preoperative computed tomography or magnetic resonance imaging studies. Fifty-three percent of the patients underwent preoperative endoscopic ultrasound with aspiration of the cystic lesion. The indication for surgical intervention was growth in the size of the cyst, symptoms, cyst size >3 cm, or presence of a mural nodule. Average cyst size was 3.1 cm.

Twelve patients underwent laparoscopic hand-assisted enucleation, and 5 patients underwent laparoscopic hand-assisted duodenal-sparing pancreatic head resection (Table 2). One patient undergoing laparoscopic hand-assisted pancreatic head resection was converted to an open procedure because of dense adhesions from prior abdominal surgery. Fourteen lesions were located in the head/uncinate of the pancreas, two were in the pancreatic body, and one was in the tail (Table 2). Patients had a 5-cm subcostal hand port and between two and four working port incisions. Eight patients underwent intraoperative ultrasound; in 7 patients the purpose was to confirm the location of the lesion, and in 1 patient it was to approximate distance from the main pancreatic duct. Eight of the patients undergoing laparoscopic hand-assisted enucleation had fibrin sealant placed on the enucleation surface. All patients had intraoperative placed drains. Final pathology was consistent with side-branch IPMN in 13 patients (1 patient with focal adenocarcinoma). Three patients had serous cysts, and 1 patient had a mucinous cyst.

Median length of stay was 6 days. There were 11 complications that occurred within 30 days postoperatively (Table 3). The most common complication was pancreatic leak. Three patients developed pancreatic leaks, which were controlled with intraoperative placed drains. Two of these patients received additional antibiotics. One patient required additional percutaneous drain placement. All four pancreatic leaks occurred in patients who underwent laparoscopic enucleation. The overall morbidity rate for our study population was 65%, although four of the complications occurred with the same patient.

It is difficult to compare the enucleation group and the duodenal-sparing pancreatic head resection group because the surgeries are very different. It is significant to note, however, all the pancreatic leaks were in the enucleation group. As expected, the length of stay was longer in the duodenal-sparing head resection group compared with the enucleation group (mean/median, 25/28 days versus 5/5 days, respectively). This was mostly secondary to delayed gastric emptying in 1 patient and multiple complications in a second patient.

Patients underwent postoperative imaging as clinically indicated and then at 1-year intervals following surgery. Median time from surgery to latest follow-up imaging is 2.8 years. No patients have developed recurrent cysts or adenocarcinoma, and all patients are currently alive.

Discussion

Multiple studies have demonstrated the safety and efficacy of the enucleation of small cystic or neuroendocrine pancreatic lesions.^10–15 Cauley et al. ¹¹ compared 45 patients who underwent enucleation of small pancreatic lesions with a matched cohort of 90 patients who underwent formal pancreatic resection. The enucleated lesions were composed of neuroendocrine tumors, mucinous cystic tumors, serous cysts, and other benign lesions. Enucleation was associated with reduced operative time, lower blood loss, and similar overall complication rates and pancreatic fistula rates compared with formal pancreatic resection. Additionally, improved endocrine and exocrine function was reported following enucleation compared with formal pancreatic resection. A multi-institutional review conducted by Pitt et al. ¹⁴ included 37 patients who underwent enucleation and 87 patients who underwent formal resection for neuroendocrine tumors of the pancreas. This review found that for tumors located in the head of the pancreas, enucleation led to reduced operative time, lower blood loss, and decreased length of stay; however, the rate of pancreatic fistula formation was significantly higher following enucleation (38% versus 15%). ¹⁴ Other studies have also reported higher pancreatic fistula rates following enucleation compared with formal pancreatic resection.^12,15 In our series of 17 patients, 4 patients developed pancreatic leaks, all of whom underwent enucleation of their lesions.

With the advent of minimally invasive techniques, centers have begun performing laparoscopic pancreatic resection and enucleation.^16–19 A systematic review included 11 studies published from 1996 to 2008, comprising a total of 101 patients who underwent laparoscopic pancreatic tumor enucleation. ¹⁶ The overall complication rate was 47%, with a 29% rate of pancreatic fistula. The authors concluded laparoscopic enucleation was a safe procedure with similar results compared with those following open enucleation. More recently, Dedieu et al. ¹⁷ described the laparoscopic enucleation of 23 patients with benign pancreatic tumors. In their report, the authors concluded the laparoscopic enucleation operative time, hospital stay, and pancreatic fistula rate seemed to be lower than those reported with open procedures.

Few studies have examined the role of pancreatic enucleation specifically for side-branch IPMN. Turrini et al. ²⁰ reported their results on 10 patients who underwent pancreatic enucleation for side-branch IPMN in the pancreatic head/uncinate compared with 100 patients who underwent pancreaticoduodenectomy for IPMN. The enucleation procedure had comparatively shorter operating times and less blood loss but similar morbidity, mortality, and length of stay in the hospital (median length of stay, 12 days). The overall number of postoperative pancreatic fistulas was higher in the enucleation group, but the majority of these fistulas were of minimal clinical significance. The authors concluded that enucleation for side-branch IPMN is a safe and acceptable procedure. Similarly, Hwang et al. ²¹ described enucleation for side-branch IPMN in 4 patients with lesions in the pancreatic head/uncinate and also reported shorter operating times and less blood loss with similar complication rates. These studies described the operative outcomes and short-term complications but did not detail any postoperative imaging follow-up. Sauvanet et al. ²² recently published a study describing 81 patients who underwent parenchymal-sparing resections for noninvasive IPMN lesions. This study included 47 branch-duct IPMN lesions treated with open enucleation or uncinate process resection. The authors reported a pancreatic fistula rate of 47% in this group of patients and a median length of stay of 21 days. The overall rate of preservation of endocrine and exocrine pancreatic function was 92%, and re-operation for recurrence was necessary in only 4% of patients.

These studies support the feasibility of laparoscopic or open enucleation for pancreatic side-branch IPMN. The possible increased risk of pancreatic fistula should be balanced against the preservation of parenchyma and pancreatic function.

On final pathology 3 of our patients had serous cystadenomas. These lesions are considered benign with a very low risk of malignancy, and therefore recommendations include resection of symptomatic lesions, lesions >4 cm, lesions with a significant growth rate, or lesions with uncertain diagnosis.^23–25 Postoperatively, with a confirmed diagnosis of benign serous cystadenoma, no postoperative surveillance is routinely performed.

There are currently no established recommendations for surveillance in patients following resection for side-branch IPMN, but clearly follow-up is necessary because of the risk of malignancy and recurrence in this group of patients.^3,26,27 The 2012 consensus guidelines suggest follow-up at 2 and 5 years in patients with negative margins and no residual lesions, as well as imaging twice yearly in patients with dysplasia at the resection margins. ² The authors noted, however, that reports from other centers recommend follow-up every 6 months because of the risk of cancer development. Sawai et al. ²⁶ followed up 103 patients with side-branch IPMN with annual endoscopic ultrasound and reported a 5-year rate of pancreatic cancer development of 2.4% and a 10-year rate of 20%, which they felt warranted continued long-term surveillance.

We have performed laparoscopic hand-assisted duodenal-sparing pancreatic head resection or enucleation in 17 patients for suspected side-branch IPMN in order to preserve pancreatic function and address the risk of malignancy associated with IPMN. In these patients, we obtain follow-up imaging on an annual basis. We have been performing this procedure for the last 5 years with a mean imaging follow-up of 2.8 years, showing no recurrence of IPMN or development of malignancy. In agreement with the aforementioned studies, our results demonstrate the safety of laparoscopic hand-assisted duodenal-sparing pancreatic head resection or enucleation for patients with side-branch IPMN, in terms of both operative outcomes and postoperative recurrence risk. Additionally, the performance of these procedures laparoscopically with hand assistance seems to lead to shorter length of stay compared with open procedures and should translate into lower overall costs.