Surg Endosc (2014) 28:1601–1606 DOI 10.1007/s00464-013-3357-x
and Other Interventional Techniques
Pure laparoscopic middle pancreatectomy: single-center experience with 13 cases Safi Dokmak • Be´atrice Aussilhou • Fadhel Samir Fte´riche Philippe Levy • Philippe Ruszniewski • Jacques Belghiti • Alain Sauvanet
•
Received: 15 August 2013 / Accepted: 26 November 2013 / Published online: 1 January 2014 Ó Springer Science+Business Media New York 2013
Abstract Background Laparoscopic pancreatic surgery is performed with increasing frequency, but laparoscopic middle pancreatectomy (LMP) is rarely described. This study aimed retrospectively to describe the authors’ unicentrically and prospectively collected data at a specialized center. Methods Since July 2011, 13 patients have undergone LMP. In this study, all their demographics and operative and postoperative data were studied from a prospectively maintained database. Results The study included eight women and five men with a mean age of 51 (range 27–75 years) and a body mass index of 26 kg/m2 (range 22–32 kg/m2). The main indications were neuroendocrine tumor (n = 7), intraductal papillary mucinous neoplasia (n = 2), solid pseudopapillary tumor (n = 2), and other (n = 2). The median duration of surgery was 190 min (range 120–285 min), and the mean blood loss was 100 ml (range 50–800 ml). Only one conversion was performed (8 %). The postoperative outcomes showed no mortality. Clinically significant pancreatic fistula (B and C) were found in 30 % of the cases. Bleeding was observed in two patients (15 %) and reintervention in three patients (23 %). The median hospital stay was 24 days (range 14–53 days), with no readmissions. The long-term follow-up evaluation showed no S. Dokmak (&) B. Aussilhou F. S. Fte´riche J. Belghiti A. Sauvanet Department of HPB Surgery and Liver Transplantation, Beaujon Hospital, Clichy, France e-mail: [email protected] P. Levy P. Ruszniewski Department of Gastroenterology, Beaujon Hospital, Clichy, France
endocrine insufficiency and only one endocrine insufficiency (8 %). Conclusions LMP is a safe surgical procedure allowing a minimally invasive approach for low malignant-potential lesions and offering a postoperative outcome comparable with that of the open approach. Keywords Laparoscopic approach Surgery Middle pancreatectomy Central pancreatectomy
As laparoscopy has progressed, laparoscopic pancreatic resections (LPRs) have been performed more frequently. Laparoscopic distal pancreatectomy (LDP) remains the most frequent procedure because it renders pancreatic anastomosis unnecessary, and the dissection is relatively safe because it is performed far from the celiomesenteric vessels. Large series have been published showing a shorter hospital stay, improvement in the number of complications [1–4], and improved oncologic safety [5]. Laparoscopic pancreaticoduodenectomy (LPD), initially described in 1994 [6], still is a challenging procedure even for experienced surgeons, especially the reconstruction phase [7]. Because this procedure is performed for selected patients, the number of cases has increased slowly [7], although the number has increased recently [8–10]. Nevertheless, it still is too early to reach conclusions on the safety and advantages of this technique compared with open approach. Reports on laparoscopic middle pancreatectomy (LMP) are rare [11–13] because this operation is considered to be difficult, even via open surgery, due to extensive dissection along the splenic vessels and the higher risk of potential postoperative pancreatic fistula at two sites [14, 15]. We believe laparoscopy is a good choice for middle
123
1602
Surg Endosc (2014) 28:1601–1606
approach. If the remnant pancreas were shorter than 5 cm, a left pancreatectomy was considered. Surgical technique
Fig. 1 Trocar placement. Five trocars usually were used for this procedure
pancreatectomy (MP) because although indications for the laparoscopic approach in LDP and LPD can be limited by oncologic or technical factors, these limitations do not exist for MP, and it usually is proposed for benign or low malignant-potential diseases with no vascular invasion. The number of MP candidates for whom the laparoscopic approach is indicated probably is fairly high. This report describes our experience in 13 patients who underwent LMP.
Patients and methods Between January 2008 and March 2013, 158 patients underwent LPR in our department of hepatobiliary and pancreatic surgery, and since July 2011, 13 patients (8 %) have undergone LMP. Prospective data were extracted from the patient database including demographics as well as surgical and postoperative outcomes. The definition of pancreatic fistula by the International Study Group of Pancreatic Diseases was used [16]. Indications The indications for LMP were similar to those for the open approach and included mainly benign and low malignantpotential diseases not suitable for enucleation. Adenocarcinoma, vascular invasion or infiltration, lesions requiring multiple frozen sections such as intraductal papillary mucinous neoplasia (IPMN), and large tumors requiring extensive lymphadenectomy were considered to be contraindications to LMP. Although extended right MP requiring right-side section in the pancreatic head were considered a contraindication to the laparoscopic approach, extended left MP was more difficult to perform but considered for the laparoscopic
123
The patient was placed in the supine position under general anesthesia with the legs spread apart and the right arm along the body. The surgeon was to the right of the patient. The assistant was between the patient’s legs, and the nurse was to the right of the surgeon (Fig. 1). Open coelioscopy was performed, and five trocars were necessary for this procedure. A 30° laparoscopic optic was inserted through a 10 mm umbilicus trocar, and a 10-mm operating trocar was inserted into the right hypochondrium. A 5- or 10-mm trocar was inserted into the left hypochondrium, and two 5-mm trocars were inserted into the right subcostal and epigastric areas for apprehension and stomach traction, respectively. A harmonic scissors and a bipolar coagulation device were needed. Once exploration was completed, the gastrocolic ligament was divided, the anterior surface of the pancreas was freed, and the stomach was charged with a tape and tracked by the epigastric trocar. The inferior pancreatic border was freed at the neck and body so the mesentericoportal vein could be identified. The superior pancreatic border was freed, and lymphadenectomy of the hepatic artery and the celiac trunk was performed along the origin of the splenic and left gastric arteries to obtain lymph node samples and facilitate dissection. Once the borders of the pancreas and the venous axis had been clearly identified, the pancreas could be easily controlled at the neck and sectioned with a stapler. We recommend right-to-left pancreatic dissection unless the right cut is located to the right of the gastroduodenal artery. In that case, a left-to-right dissection is recommended, allowing better preparation of the right transection line. The middle pancreas was freed from the splenic vessels. Ligation and division of small collaterals were facilitated by a harmonic scissors and small clips. The pancreas then was sectioned on the left (to the left of the lesion) with a stapler or harmonic scissors. The main pancreatic duct was repermeabilized after partial resection of the stapler line. The specimen, which usually was small, could be retrieved via the trocar site, which was enlarged to 2–3 cm. Onelayer pancreaticogastric anastomosis was completed with interrupted sutures of Vicryl 3/0. The right side of the pancreas and the anastomosis were drained with tubular drains exteriorized by the trocar sites. Location of the tumor Perioperative localization of the tumor can be difficult, especially if it is deep. Although perioperative ultrasound
Surg Endosc (2014) 28:1601–1606
1603 Table 1 Postoperative complications
Fig. 2 Computed tomography (CT) scan of a patient with branchduct IPMN. Surgical landmarks are to the right. The gastroduodenal artery (blue arrow), the portal vein (black arrow), and the curve of the splenic artery are to the left (white arrow) (Color figure online)
Variables
n (%)
Mortality
0 (0)
Global morbidity
10/13 (77)
Pancreatic fistula
9 (69)
Grade A
4 (31)
Grade B
2 (15)
Grade C
2 (15)
Peripancreatic collections
3 (23)
Bleeding
2 (15)
Biliary fistula
1 (8)
Reintervention
3 (23)
Delayed gastric emptying Pulmonary complication
1 (8) 1 (8)
Readmission
0 (0)
Median hospital stay: days range)
24 (14–53)
has been used for this purpose, we based tumor localization mainly on preoperative imaging. The main landmarks are the gastroduodenal artery and the mesentericoportal vein to the right and the first curve of the splenic artery to the left (Fig. 2). Although this method of tumor localization is not particularly helpful for lesions located in the neck, it can be quite pertinent for deeply located lesions in the body of the pancreas (left-side MP).
noted that one patient underwent LMP for main duct stenosis, shown by acute pancreatitis, with no apparent mass on imaging and a negative frozen section. This patient underwent left splenopancreatectomy 1 week later because the definitive histology results of the LMP specimen were positive for adenocarcinoma. Two patients underwent LMP extended to the right along the gastroduodenal artery. Two patients had leftsided LMP, and one patient underwent LMP with ligation of the splenic artery.
Results
Postoperative outcome
During the study period, 13 patients (8 women and 5 men) underwent LMP. The mean age of these patients was 51 years (range 27–75 years), with 4 patients younger than 40 years. The mean body mass index (BMI) was 26 kg/m2 (range 22–32 kg/m2), and four of the patients had a BMI higher than 30 kg/m2. The main indications were neuroendocrine tumor (NET) (n = 7), IPMN (n = 2), solid pseudopapillary tumor (SPPT; n = 2), mucinous cystadenoma (n = 1), and main pancreatic duct stenosis, which was shown to be adenocarcinoma (n = 1).
The postoperative complications are summarized in Table 1. No mortalities occurred, and the overall morbidity rate was 77 % (10/13). Clinically significant pancreatic fistulas (B and C) were observed in 4 of the 13 patients (31 %). Pancreatic fistula was complicated by symptomatic collections in two patients, respectively requiring percutaneous drainage (n = 1) and surgical treatment (n = 1). Although in some cases it can be difficult to determine the origin of pancreatic fistula (anastomosis or right pancreatic stump?), according to our clinical and radiologic data, most pancreatic fistulas were related to the pancreatic anastomosis. Postoperative bleeding observed in two patients (15 %) was treated by embolization of the splenic artery (n = 1) in the patient who received platelet anti-aggregant treatment and by ligation of the gastroduodenal artery (n = 1) in the patient who underwent extended LMP to the right. Biliary fistula from the cystic stump was observed in one patient who underwent an associated cholecystectomy. Three patients (23 %) underwent reintervention (on postoperative days 14, 18, and 20, respectively) by the laparoscopic approach due to small bowel occlusion caused by drainage
Operative data The median duration of surgery was 190 min (range 120–285 min) for the entire population, 236 min (range 180–285 min) for the first six patients, and 143 min (range 130–240 min) for the last seven patients (p \ 0.001). The mean blood loss was 100 ml (range 50–800 min), and only one converted patient received a transfusion. Pure laparoscopy was performed in all cases, with one conversion (8 %) performed for bleeding. It should be
123
1604
adhesions (n = 1), and by laparotomy for bleeding from the gastroduodenal artery (n = 1), and peripancreatic collection (n = 1). The median hospital stay was 24 days (range 14–53 days), and no patient had readmission. Pathology The final pathology confirmed the preoperative diagnosis for all except one patient (92 %, 12/13) who had main duct stenosis without a mass, which was identified as adenocarcinoma. The median tumor size of the solid lesions was 1.8 cm (range 0.7–4.2 cm). The median specimen length was 5 cm (range 4–7 cm). For the all population, the median number of retrieved lymph nodes (LN) was 3 (range 0–6). Four patients (1 with conversion) operated for IPMN (n = 1), SPPT (n = 1) and NETs (n = 2) had no LN on the resected specimen. All the LNs were negative except in one patient with adenocarcinoma, who had 3 positive LN among the 4 retrieved with the specimen. The resection was R0 in all cases (100 %). Follow-up evaluation No tumor recurrence was observed after a mean follow-up period of 9 months (range 1–19 months). No clinical pancreatic exocrine insufficiency was noted, and only 1 (8 %) of the 13 patients had upper limit glycemia. Two patients had asymptomatic incisional hernia. The hernia was at the trocar site in one patient and on the midline incision performed for reintervention in the other patient. One patient had stenosis of the pancreaticogastric anastomosis, with pancreatic gland atrophy.
Discussion Due to widespread use of imaging techniques, benign and low malignant-potential diseases are more frequently discovered. In the twenty first century, these lesions should be treated by parenchymal-preserving resection (MP) and by a mini-invasive approach. If enucleation is contraindicated, MP represents an attractive surgical alternative to left pancreatectomy or pancreaticoduodenectomy, especially when pancreatic lesions are more to the left or right of the pancreatic neck. Although MP is thought to be associated with increased postoperative morbidity and mortality due to a higher risk of pancreatic fistula from two potential pancreatic stumps (right pancreatic stump and anastomosis), this procedure has an excellent long-term outcome for the risk of endocrine or exocrine insufficiency compared with other pancreatic surgical techniques [17]. These points together with abdominal wall integrity and the cosmetic results of surgery are very important, especially when
123
Surg Endosc (2014) 28:1601–1606
benign or low malignant-potential diseases are being treated in relatively young patients. Open MP is a well-established pancreatic resection technique described in several published series [14, 18–22]. In the largest series ([50 patients) [14, 18–20], as in any retrospective study, it is very difficult to interpret the postoperative outcome because the results are heterogeneous, particularly in the definition of pancreatic fistula and endocrine and exocrine insufficiency. These studies have shown a mortality rate of 0–2.5 %, an overall morbidity rate of 41–58 %, a pancreatic fistula rate of 8–44 %, an endocrine insufficiency rate of 0–14 %, and an exocrine insufficiency rate of 7–22 %. A recent metaanalysis comparing central pancreatectomy (CP) with distal pancreatectomy (DP) showed CP to be a safe procedure, carrying a higher risk of morbidity but very good long-term functional results compared to DP [17]. These results show that mortality is not nil and morbidity is comparable with that of other major pancreatic procedures (PD and LP) or slightly higher, but that the results are very good for long-term endocrine and exocrine insufficiency. Although LMP has been described in small series in the literature [11–13] and in case reports [23–25], only three studies [11–13] have reported on more than five patients (Table 2). In these series, the main indications for resection were serous cystadenoma (n = 7), NET (n = 6), solid pseudopapillay tumors (n = 4), focal pancreatitis (n = 1), mucinous cystadenoma (n = 1), and IPMN (n = 1). The median duration of surgery was 380 min (range 225–480 min). The conversion rate varied between 17 and 22 %. No mortality occurred, and the morbidity rate was 20–45 %. The median reported hospital stay was 12 days (range 3–41 days). Although serous cystadenomas are the main indication for surgical resection in the literature, all our patients had low malignant-potential diseases, and none had serous cystadenoma. This was due to our complete radiologic and endoscopic assessment for all pancreatic cystic lesions to eliminate serous cystadenoma, which still are a contraindication to surgical resection in our practice. Moreover, our experience with the laparoscopic approach does not extend our surgical resections to include benign lesions. The reduced duration of surgery in this series compared with the literature (200 vs 385 min) could have been due to our experience with this technique by both the open and laparoscopic approaches, our experience with laparoscopic surgery ([150 cases including 30 with pancreaticoduodenectomy and 2 with total pancreatectomy), and our strategy of using pancreaticogastric and not pancreaticojejunal anastomosis, which is more time consuming. The only patient in our series who underwent conversion had a large NET extending to the right behind the gastroduodenal artery, which initially limits the indication for the laparoscopic approach. In this study, LMP was associated with higher morbidity (77 %), clinically significant pancreatic fistula (30 %), and
Surg Endosc (2014) 28:1601–1606
1605
Table 2 Largest reported series in the literature on central or middle laparoscopic pancreatectomy Author [ref], journal, date
Period
n
Conversion n (%)
Blood loss ml (range)
Duration min (range)
Morbidity (%)
Hospital stay days (range)
Sa Cunha et al. [11], Surgery, 2007
1999–2006
6
1/6 (17)
125 (50–300)
225 (180–365)
33
18 (15–25)
2/9 (22)
Rotellar et al. [12], Ann Surg, 2008
2005–2007
9
Kang et al. [13], Surg Endosc, 2011
2007–2009
5
Current series
2011–2013
13
1/13 (8)
additional surgery (23 %). However, these results were similar to those of our open approach, in which LMP also was associated with overall morbidity (72 %), clinically significant pancreatic fistula (44 %), and mortality (3 %) (nonpublished data). This significant morbidity was related mainly to grade A pancreatic fistula. No deaths occurred, and the rates for clinically significant pancreatic fistula (grades B and C) were similar to those with the open approach [17]. The patients in whom grade C fistula developed had undergone either extended right LMP or were receiving anticoagulation treatment. Our data suggest that most pancreatic fistulas were related to pancreaticogastric anastomosis, and we believe that a subgroup of patients may benefit more from distal pancreatectomy than MP to avoid pancreatic anastomosis. This subgroup includes elderly patients, patients with associated comorbidities, patients receiving anticoagulation treatment, and patients with previous diabetes, especially if the texture of the pancreas is soft. In retrospect, the 65-year-old patient in our study with vascular disease and anticoagulation treatment probably should have undergone distal pancreatectomy because of the complicated outcome (fistula, bleeding, and reintervention). MP probably should be limited to young patients without severe comorbidities. At the same time and as with any new surgical procedure, a learning curve exists, and it is too early to compare our results with a second cohort because of the small study population. The hospital stay was longer than reported in the literature for the open approach (17 days; range 9–65 days) [17] or the laparoscopic approach (12 days; range 3–41 days). This could have resulted from our postoperative strategy, whereby patients received nothing per mouth for at least 7 days, all fistulas were managed in the hospital, and no patient was discharged before pancreatic fistula had completely healed. This strategy also explains the low rate of readmission and of patients requiring percutaneous drainage for collections. The resection was R0 in all cases, and in 9 (70 %) of the 13 of cases, a mean number of 3 LNs (range 0–6 LNs) could be obtained, which allowed node-positive tumors to be identified and complementary resection to be performed if necessary, as for our patient with adenocarcinoma. Widespread use of LPR is limited by oncologic and technical factors (obesity, inflammation, and anastomoses). Compared with other LPRs, LMP can benefit most from
?
435 (357–509)
45
5 (3–41)
200 (100–600)
480 (360–480)
20
12 (9–28)
100 (50–800)
190 (120–285)
Global 77 Major 30
24 (14–53)
the laparoscopic approach because it generally is proposed for benign or low malignant-potential disease, and thus is without oncologic limitations. Technically, MP has an advantage over other pancreatic resection techniques. For example, in case of MP, obesity is not a major technical problem because the relatively superficial location of the central pancreas makes its dissection relatively easy compared with the use of pancreaticoduodenectomy or splenopancreatectomy, and of course, only one anastomosis is needed. The only limiting factor for LMP is inflammatory adhesions or lesions necessitating extended LMP. This good indication for the laparoscopic approach can be illustrated by the fact that since we started our experience with LMP, all candidates for MP except one have undergone surgery by the laparoscopic approach (13/14, 93 %).
Conclusion Laparoscopic MP is a safe surgical procedure that has a postoperative outcome comparable with that of open surgery. This approach has the advantage of allowing resection of benign or low malignant-potential disease by a minimally invasive approach. This technique should be learned because with experienced surgeons, most candidates for MP should theoretically benefit from the laparoscopic approach.
Disclosures Safi Dokmak, Be´atrice Aussilhou, Fadhel Samir Fte´riche, Philippe Levy, Philippe Ruszniewski, Jacques Belghiti, and Alain Sauvanet have no conflicts of interest or financial ties to disclose.
References 1. Kooby DA, Gillespie T, Bentrem DJ et al (2008) Left-sided pancreatectomy: a multicentre comparison of laparoscopic and open approaches. Ann Surg 248:438–446 2. Røsok BI, Marangos IP, Kazaryan AM, Rosseland AR, Buanes T et al (2010) Single-centre experience of laparoscopic pancreatic surgery. Br J Surg 97:902–909. doi:10.1002/bjs.7020 3. Song KB, Kim SC, Park JB, Kim YH, Jung YS, Kim MH et al (2011) Single-center experience of laparoscopic left pancreatic
123
1606
4.
5.
6. 7.
8.
9.
10.
11.
12.
13.
14.
resection in 359 consecutive patients: changing the surgical paradigm of left pancreatic resection. Surg Endosc 25:3364–3372 Venkat R, Edil BH, Schulick RD, Lidor AO, Makary MA, Wolfgang CL (2012) Laparoscopic distal pancreatectomy is associated with significantly less overall morbidity compared to the open technique: a systematic review and meta-analysis. Ann Surg 255:1048–1059 Kooby DA, Hawkins WG, Schmidt CM, Weber SM, Bentrem DJ, Gillespie TW et al (2010) A multicenter analysis of distal pancreatectomy for adenocarcinoma: is laparoscopic resection appropriate? J Am Coll Surg 210:779–785, 786–787 Gagner M, Pomp A (1994) Laparoscopic pylorus-preserving pancreatoduodenectomy. Surg Endosc 8:408–410 Gagner M, Palermo M (2009) Laparoscopic Whipple procedure: review of the literature. J Hepatobiliary Pancreat Surg 16: 726–730 Epub 28 July 2009 Kendrick ML, Cusati D (2010) Total laparoscopic pancreaticoduodenectomy: feasibility and outcome in an early experience. Arch Surg 145:19–23 Zeh HJ, Zureikat AH, Secrest A, Dauoudi M, Bartlett D, Moser AJ (2012) Outcomes after robot-assisted pancreaticoduodenectomy for periampullary lesions. Ann Surg Oncol 19:864–870 Asbun HJ, Stauffer JA (2012) Laparoscopic vs open pancreaticoduodenectomy: overall outcomes and severity of complications using the accordion severity grading system. J Am Coll Surg 215:810–819 Sa Cunha A, Rault A, Beau C, Collet D, Masson B (2007) Laparoscopic central pancreatectomy: single-institution experience of 6 patients. Surgery 142:405–409 Rotellar F, Pardo F, Montiel C, Benito A, Regueira FM, Poveda I et al (2008) Totally laparoscopic Roux-en-Y duct-to-mucosa pancreaticojejunostomy after middle pancreatectomy: a consecutive nine-case series at a single institution. Ann Surg 247:938–944 Kang CM, Kim DH, Lee WJ, Chi HS (2011) Initial experiences using robot-assisted central pancreatectomy with pancreaticogastrostomy: a potential way to advanced laparoscopic pancreatectomy. Surg Endosc 25:1101–1106. doi:10.1007/s00464-0101324-3 Sauvanet A, Partensky C, Sastre B, Gigot JF, Fagniez PL, Tuech JJ et al (2002) Medial pancreatectomy: a multi-institutional retrospective study of 53 patients by the French Pancreas Club. Surgery 132:836–843
123
Surg Endosc (2014) 28:1601–1606 15. Christein JD, Smoot RL, Farnell MB (2006) Central pancreatectomy: a technique for the resection of pancreatic neck lesions. Arch Surg 141:293–299 16. Bassi C, Dervenis C, Butturini G, Fingerhut A, Yeo C, Izbicki J et al (2005) Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery 138:8–13 17. Iacono C, Verlato G, Ruzzenente A, Campagnaro T, Bacchelli C, Valdegamberi A et al (2013) Systematic review of central pancreatectomy and meta-analysis of central versus distal pancreatectomy. Br J Surg 100:873–885. doi:10.1002/bjs.9136 18. Crippa S, Bassi C, Warshaw AL, Falconi M, Partelli S, Thayer SP et al (2007) Middle pancreatectomy: indications, short- and longterm operative outcomes. Ann Surg 246:69–76 19. DiNorcia J, Ahmed L, Lee MK, Reavey PL, Yakaitis EA, Lee JA et al (2010) Better preservation of endocrine function after central versus distal pancreatectomy for mid-gland lesions. Surgery 148:1247–1254. doi:10.1016/j.surg.2010.09.003 (discussion 1254–1256) 20. Adham M, Giunippero A, Hervieu V, Courbie`re M, Partensky C (2008) Central pancreatectomy: single-center experience of 50 cases. Arch Surg 143:175–180. doi:10.1001/archsurg.2007.52 (discussion 180–181) 21. Shikano T, Nakao A, Kodera Y, Yamada S, Fujii T, Sugimoto H et al (2010) Middle pancreatectomy: safety and long-term results. Surgery 147:21–29. doi:10.1016/j.surg.2009.04.036 22. Xiang GM, Tan CL, Zhang H, Ran X, Mai G, Liu XB (2012) Central pancreatectomy for benign or borderline lesions of the pancreatic neck: a single-centre experience and literature review. Hepatogastroenterology 59:1286–1289. doi:10.5754/hge11937 23. Kitasato A, Adachi T, Inokuma T, Tajima Y, Kanematsu T, Kuroki T (2012) Laparoscopic middle pancreatectomy under a pancreatic duct-navigation surgery. Hepatogastroenterology 59(120):2400 24. Kim DH, Kang CM, Lee WJ, Chi HS (2011) Robotique central pancreatectomy with pancreaticogastrostomy (transgastric approach) in a solid pseudopapillary tumor of the pancreas. Hepatogastroenterology 58:1805–1808 25. Giulianotti PC, Sbrana F, Bianco FM, Addeo P, Caravaglios G (2010) Robot-assisted laparoscopic middle pancreatectomy. J Laparoendosc Adv Surg Tech A 20:135–139
and Other Interventional Techniques
Pure laparoscopic middle pancreatectomy: single-center experience with 13 cases Safi Dokmak • Be´atrice Aussilhou • Fadhel Samir Fte´riche Philippe Levy • Philippe Ruszniewski • Jacques Belghiti • Alain Sauvanet
•
Received: 15 August 2013 / Accepted: 26 November 2013 / Published online: 1 January 2014 Ó Springer Science+Business Media New York 2013
Abstract Background Laparoscopic pancreatic surgery is performed with increasing frequency, but laparoscopic middle pancreatectomy (LMP) is rarely described. This study aimed retrospectively to describe the authors’ unicentrically and prospectively collected data at a specialized center. Methods Since July 2011, 13 patients have undergone LMP. In this study, all their demographics and operative and postoperative data were studied from a prospectively maintained database. Results The study included eight women and five men with a mean age of 51 (range 27–75 years) and a body mass index of 26 kg/m2 (range 22–32 kg/m2). The main indications were neuroendocrine tumor (n = 7), intraductal papillary mucinous neoplasia (n = 2), solid pseudopapillary tumor (n = 2), and other (n = 2). The median duration of surgery was 190 min (range 120–285 min), and the mean blood loss was 100 ml (range 50–800 ml). Only one conversion was performed (8 %). The postoperative outcomes showed no mortality. Clinically significant pancreatic fistula (B and C) were found in 30 % of the cases. Bleeding was observed in two patients (15 %) and reintervention in three patients (23 %). The median hospital stay was 24 days (range 14–53 days), with no readmissions. The long-term follow-up evaluation showed no S. Dokmak (&) B. Aussilhou F. S. Fte´riche J. Belghiti A. Sauvanet Department of HPB Surgery and Liver Transplantation, Beaujon Hospital, Clichy, France e-mail: [email protected] P. Levy P. Ruszniewski Department of Gastroenterology, Beaujon Hospital, Clichy, France
endocrine insufficiency and only one endocrine insufficiency (8 %). Conclusions LMP is a safe surgical procedure allowing a minimally invasive approach for low malignant-potential lesions and offering a postoperative outcome comparable with that of the open approach. Keywords Laparoscopic approach Surgery Middle pancreatectomy Central pancreatectomy
As laparoscopy has progressed, laparoscopic pancreatic resections (LPRs) have been performed more frequently. Laparoscopic distal pancreatectomy (LDP) remains the most frequent procedure because it renders pancreatic anastomosis unnecessary, and the dissection is relatively safe because it is performed far from the celiomesenteric vessels. Large series have been published showing a shorter hospital stay, improvement in the number of complications [1–4], and improved oncologic safety [5]. Laparoscopic pancreaticoduodenectomy (LPD), initially described in 1994 [6], still is a challenging procedure even for experienced surgeons, especially the reconstruction phase [7]. Because this procedure is performed for selected patients, the number of cases has increased slowly [7], although the number has increased recently [8–10]. Nevertheless, it still is too early to reach conclusions on the safety and advantages of this technique compared with open approach. Reports on laparoscopic middle pancreatectomy (LMP) are rare [11–13] because this operation is considered to be difficult, even via open surgery, due to extensive dissection along the splenic vessels and the higher risk of potential postoperative pancreatic fistula at two sites [14, 15]. We believe laparoscopy is a good choice for middle
123
1602
Surg Endosc (2014) 28:1601–1606
approach. If the remnant pancreas were shorter than 5 cm, a left pancreatectomy was considered. Surgical technique
Fig. 1 Trocar placement. Five trocars usually were used for this procedure
pancreatectomy (MP) because although indications for the laparoscopic approach in LDP and LPD can be limited by oncologic or technical factors, these limitations do not exist for MP, and it usually is proposed for benign or low malignant-potential diseases with no vascular invasion. The number of MP candidates for whom the laparoscopic approach is indicated probably is fairly high. This report describes our experience in 13 patients who underwent LMP.
Patients and methods Between January 2008 and March 2013, 158 patients underwent LPR in our department of hepatobiliary and pancreatic surgery, and since July 2011, 13 patients (8 %) have undergone LMP. Prospective data were extracted from the patient database including demographics as well as surgical and postoperative outcomes. The definition of pancreatic fistula by the International Study Group of Pancreatic Diseases was used [16]. Indications The indications for LMP were similar to those for the open approach and included mainly benign and low malignantpotential diseases not suitable for enucleation. Adenocarcinoma, vascular invasion or infiltration, lesions requiring multiple frozen sections such as intraductal papillary mucinous neoplasia (IPMN), and large tumors requiring extensive lymphadenectomy were considered to be contraindications to LMP. Although extended right MP requiring right-side section in the pancreatic head were considered a contraindication to the laparoscopic approach, extended left MP was more difficult to perform but considered for the laparoscopic
123
The patient was placed in the supine position under general anesthesia with the legs spread apart and the right arm along the body. The surgeon was to the right of the patient. The assistant was between the patient’s legs, and the nurse was to the right of the surgeon (Fig. 1). Open coelioscopy was performed, and five trocars were necessary for this procedure. A 30° laparoscopic optic was inserted through a 10 mm umbilicus trocar, and a 10-mm operating trocar was inserted into the right hypochondrium. A 5- or 10-mm trocar was inserted into the left hypochondrium, and two 5-mm trocars were inserted into the right subcostal and epigastric areas for apprehension and stomach traction, respectively. A harmonic scissors and a bipolar coagulation device were needed. Once exploration was completed, the gastrocolic ligament was divided, the anterior surface of the pancreas was freed, and the stomach was charged with a tape and tracked by the epigastric trocar. The inferior pancreatic border was freed at the neck and body so the mesentericoportal vein could be identified. The superior pancreatic border was freed, and lymphadenectomy of the hepatic artery and the celiac trunk was performed along the origin of the splenic and left gastric arteries to obtain lymph node samples and facilitate dissection. Once the borders of the pancreas and the venous axis had been clearly identified, the pancreas could be easily controlled at the neck and sectioned with a stapler. We recommend right-to-left pancreatic dissection unless the right cut is located to the right of the gastroduodenal artery. In that case, a left-to-right dissection is recommended, allowing better preparation of the right transection line. The middle pancreas was freed from the splenic vessels. Ligation and division of small collaterals were facilitated by a harmonic scissors and small clips. The pancreas then was sectioned on the left (to the left of the lesion) with a stapler or harmonic scissors. The main pancreatic duct was repermeabilized after partial resection of the stapler line. The specimen, which usually was small, could be retrieved via the trocar site, which was enlarged to 2–3 cm. Onelayer pancreaticogastric anastomosis was completed with interrupted sutures of Vicryl 3/0. The right side of the pancreas and the anastomosis were drained with tubular drains exteriorized by the trocar sites. Location of the tumor Perioperative localization of the tumor can be difficult, especially if it is deep. Although perioperative ultrasound
Surg Endosc (2014) 28:1601–1606
1603 Table 1 Postoperative complications
Fig. 2 Computed tomography (CT) scan of a patient with branchduct IPMN. Surgical landmarks are to the right. The gastroduodenal artery (blue arrow), the portal vein (black arrow), and the curve of the splenic artery are to the left (white arrow) (Color figure online)
Variables
n (%)
Mortality
0 (0)
Global morbidity
10/13 (77)
Pancreatic fistula
9 (69)
Grade A
4 (31)
Grade B
2 (15)
Grade C
2 (15)
Peripancreatic collections
3 (23)
Bleeding
2 (15)
Biliary fistula
1 (8)
Reintervention
3 (23)
Delayed gastric emptying Pulmonary complication
1 (8) 1 (8)
Readmission
0 (0)
Median hospital stay: days range)
24 (14–53)
has been used for this purpose, we based tumor localization mainly on preoperative imaging. The main landmarks are the gastroduodenal artery and the mesentericoportal vein to the right and the first curve of the splenic artery to the left (Fig. 2). Although this method of tumor localization is not particularly helpful for lesions located in the neck, it can be quite pertinent for deeply located lesions in the body of the pancreas (left-side MP).
noted that one patient underwent LMP for main duct stenosis, shown by acute pancreatitis, with no apparent mass on imaging and a negative frozen section. This patient underwent left splenopancreatectomy 1 week later because the definitive histology results of the LMP specimen were positive for adenocarcinoma. Two patients underwent LMP extended to the right along the gastroduodenal artery. Two patients had leftsided LMP, and one patient underwent LMP with ligation of the splenic artery.
Results
Postoperative outcome
During the study period, 13 patients (8 women and 5 men) underwent LMP. The mean age of these patients was 51 years (range 27–75 years), with 4 patients younger than 40 years. The mean body mass index (BMI) was 26 kg/m2 (range 22–32 kg/m2), and four of the patients had a BMI higher than 30 kg/m2. The main indications were neuroendocrine tumor (NET) (n = 7), IPMN (n = 2), solid pseudopapillary tumor (SPPT; n = 2), mucinous cystadenoma (n = 1), and main pancreatic duct stenosis, which was shown to be adenocarcinoma (n = 1).
The postoperative complications are summarized in Table 1. No mortalities occurred, and the overall morbidity rate was 77 % (10/13). Clinically significant pancreatic fistulas (B and C) were observed in 4 of the 13 patients (31 %). Pancreatic fistula was complicated by symptomatic collections in two patients, respectively requiring percutaneous drainage (n = 1) and surgical treatment (n = 1). Although in some cases it can be difficult to determine the origin of pancreatic fistula (anastomosis or right pancreatic stump?), according to our clinical and radiologic data, most pancreatic fistulas were related to the pancreatic anastomosis. Postoperative bleeding observed in two patients (15 %) was treated by embolization of the splenic artery (n = 1) in the patient who received platelet anti-aggregant treatment and by ligation of the gastroduodenal artery (n = 1) in the patient who underwent extended LMP to the right. Biliary fistula from the cystic stump was observed in one patient who underwent an associated cholecystectomy. Three patients (23 %) underwent reintervention (on postoperative days 14, 18, and 20, respectively) by the laparoscopic approach due to small bowel occlusion caused by drainage
Operative data The median duration of surgery was 190 min (range 120–285 min) for the entire population, 236 min (range 180–285 min) for the first six patients, and 143 min (range 130–240 min) for the last seven patients (p \ 0.001). The mean blood loss was 100 ml (range 50–800 min), and only one converted patient received a transfusion. Pure laparoscopy was performed in all cases, with one conversion (8 %) performed for bleeding. It should be
123
1604
adhesions (n = 1), and by laparotomy for bleeding from the gastroduodenal artery (n = 1), and peripancreatic collection (n = 1). The median hospital stay was 24 days (range 14–53 days), and no patient had readmission. Pathology The final pathology confirmed the preoperative diagnosis for all except one patient (92 %, 12/13) who had main duct stenosis without a mass, which was identified as adenocarcinoma. The median tumor size of the solid lesions was 1.8 cm (range 0.7–4.2 cm). The median specimen length was 5 cm (range 4–7 cm). For the all population, the median number of retrieved lymph nodes (LN) was 3 (range 0–6). Four patients (1 with conversion) operated for IPMN (n = 1), SPPT (n = 1) and NETs (n = 2) had no LN on the resected specimen. All the LNs were negative except in one patient with adenocarcinoma, who had 3 positive LN among the 4 retrieved with the specimen. The resection was R0 in all cases (100 %). Follow-up evaluation No tumor recurrence was observed after a mean follow-up period of 9 months (range 1–19 months). No clinical pancreatic exocrine insufficiency was noted, and only 1 (8 %) of the 13 patients had upper limit glycemia. Two patients had asymptomatic incisional hernia. The hernia was at the trocar site in one patient and on the midline incision performed for reintervention in the other patient. One patient had stenosis of the pancreaticogastric anastomosis, with pancreatic gland atrophy.
Discussion Due to widespread use of imaging techniques, benign and low malignant-potential diseases are more frequently discovered. In the twenty first century, these lesions should be treated by parenchymal-preserving resection (MP) and by a mini-invasive approach. If enucleation is contraindicated, MP represents an attractive surgical alternative to left pancreatectomy or pancreaticoduodenectomy, especially when pancreatic lesions are more to the left or right of the pancreatic neck. Although MP is thought to be associated with increased postoperative morbidity and mortality due to a higher risk of pancreatic fistula from two potential pancreatic stumps (right pancreatic stump and anastomosis), this procedure has an excellent long-term outcome for the risk of endocrine or exocrine insufficiency compared with other pancreatic surgical techniques [17]. These points together with abdominal wall integrity and the cosmetic results of surgery are very important, especially when
123
Surg Endosc (2014) 28:1601–1606
benign or low malignant-potential diseases are being treated in relatively young patients. Open MP is a well-established pancreatic resection technique described in several published series [14, 18–22]. In the largest series ([50 patients) [14, 18–20], as in any retrospective study, it is very difficult to interpret the postoperative outcome because the results are heterogeneous, particularly in the definition of pancreatic fistula and endocrine and exocrine insufficiency. These studies have shown a mortality rate of 0–2.5 %, an overall morbidity rate of 41–58 %, a pancreatic fistula rate of 8–44 %, an endocrine insufficiency rate of 0–14 %, and an exocrine insufficiency rate of 7–22 %. A recent metaanalysis comparing central pancreatectomy (CP) with distal pancreatectomy (DP) showed CP to be a safe procedure, carrying a higher risk of morbidity but very good long-term functional results compared to DP [17]. These results show that mortality is not nil and morbidity is comparable with that of other major pancreatic procedures (PD and LP) or slightly higher, but that the results are very good for long-term endocrine and exocrine insufficiency. Although LMP has been described in small series in the literature [11–13] and in case reports [23–25], only three studies [11–13] have reported on more than five patients (Table 2). In these series, the main indications for resection were serous cystadenoma (n = 7), NET (n = 6), solid pseudopapillay tumors (n = 4), focal pancreatitis (n = 1), mucinous cystadenoma (n = 1), and IPMN (n = 1). The median duration of surgery was 380 min (range 225–480 min). The conversion rate varied between 17 and 22 %. No mortality occurred, and the morbidity rate was 20–45 %. The median reported hospital stay was 12 days (range 3–41 days). Although serous cystadenomas are the main indication for surgical resection in the literature, all our patients had low malignant-potential diseases, and none had serous cystadenoma. This was due to our complete radiologic and endoscopic assessment for all pancreatic cystic lesions to eliminate serous cystadenoma, which still are a contraindication to surgical resection in our practice. Moreover, our experience with the laparoscopic approach does not extend our surgical resections to include benign lesions. The reduced duration of surgery in this series compared with the literature (200 vs 385 min) could have been due to our experience with this technique by both the open and laparoscopic approaches, our experience with laparoscopic surgery ([150 cases including 30 with pancreaticoduodenectomy and 2 with total pancreatectomy), and our strategy of using pancreaticogastric and not pancreaticojejunal anastomosis, which is more time consuming. The only patient in our series who underwent conversion had a large NET extending to the right behind the gastroduodenal artery, which initially limits the indication for the laparoscopic approach. In this study, LMP was associated with higher morbidity (77 %), clinically significant pancreatic fistula (30 %), and
Surg Endosc (2014) 28:1601–1606
1605
Table 2 Largest reported series in the literature on central or middle laparoscopic pancreatectomy Author [ref], journal, date
Period
n
Conversion n (%)
Blood loss ml (range)
Duration min (range)
Morbidity (%)
Hospital stay days (range)
Sa Cunha et al. [11], Surgery, 2007
1999–2006
6
1/6 (17)
125 (50–300)
225 (180–365)
33
18 (15–25)
2/9 (22)
Rotellar et al. [12], Ann Surg, 2008
2005–2007
9
Kang et al. [13], Surg Endosc, 2011
2007–2009
5
Current series
2011–2013
13
1/13 (8)
additional surgery (23 %). However, these results were similar to those of our open approach, in which LMP also was associated with overall morbidity (72 %), clinically significant pancreatic fistula (44 %), and mortality (3 %) (nonpublished data). This significant morbidity was related mainly to grade A pancreatic fistula. No deaths occurred, and the rates for clinically significant pancreatic fistula (grades B and C) were similar to those with the open approach [17]. The patients in whom grade C fistula developed had undergone either extended right LMP or were receiving anticoagulation treatment. Our data suggest that most pancreatic fistulas were related to pancreaticogastric anastomosis, and we believe that a subgroup of patients may benefit more from distal pancreatectomy than MP to avoid pancreatic anastomosis. This subgroup includes elderly patients, patients with associated comorbidities, patients receiving anticoagulation treatment, and patients with previous diabetes, especially if the texture of the pancreas is soft. In retrospect, the 65-year-old patient in our study with vascular disease and anticoagulation treatment probably should have undergone distal pancreatectomy because of the complicated outcome (fistula, bleeding, and reintervention). MP probably should be limited to young patients without severe comorbidities. At the same time and as with any new surgical procedure, a learning curve exists, and it is too early to compare our results with a second cohort because of the small study population. The hospital stay was longer than reported in the literature for the open approach (17 days; range 9–65 days) [17] or the laparoscopic approach (12 days; range 3–41 days). This could have resulted from our postoperative strategy, whereby patients received nothing per mouth for at least 7 days, all fistulas were managed in the hospital, and no patient was discharged before pancreatic fistula had completely healed. This strategy also explains the low rate of readmission and of patients requiring percutaneous drainage for collections. The resection was R0 in all cases, and in 9 (70 %) of the 13 of cases, a mean number of 3 LNs (range 0–6 LNs) could be obtained, which allowed node-positive tumors to be identified and complementary resection to be performed if necessary, as for our patient with adenocarcinoma. Widespread use of LPR is limited by oncologic and technical factors (obesity, inflammation, and anastomoses). Compared with other LPRs, LMP can benefit most from
?
435 (357–509)
45
5 (3–41)
200 (100–600)
480 (360–480)
20
12 (9–28)
100 (50–800)
190 (120–285)
Global 77 Major 30
24 (14–53)
the laparoscopic approach because it generally is proposed for benign or low malignant-potential disease, and thus is without oncologic limitations. Technically, MP has an advantage over other pancreatic resection techniques. For example, in case of MP, obesity is not a major technical problem because the relatively superficial location of the central pancreas makes its dissection relatively easy compared with the use of pancreaticoduodenectomy or splenopancreatectomy, and of course, only one anastomosis is needed. The only limiting factor for LMP is inflammatory adhesions or lesions necessitating extended LMP. This good indication for the laparoscopic approach can be illustrated by the fact that since we started our experience with LMP, all candidates for MP except one have undergone surgery by the laparoscopic approach (13/14, 93 %).
Conclusion Laparoscopic MP is a safe surgical procedure that has a postoperative outcome comparable with that of open surgery. This approach has the advantage of allowing resection of benign or low malignant-potential disease by a minimally invasive approach. This technique should be learned because with experienced surgeons, most candidates for MP should theoretically benefit from the laparoscopic approach.
Disclosures Safi Dokmak, Be´atrice Aussilhou, Fadhel Samir Fte´riche, Philippe Levy, Philippe Ruszniewski, Jacques Belghiti, and Alain Sauvanet have no conflicts of interest or financial ties to disclose.
References 1. Kooby DA, Gillespie T, Bentrem DJ et al (2008) Left-sided pancreatectomy: a multicentre comparison of laparoscopic and open approaches. Ann Surg 248:438–446 2. Røsok BI, Marangos IP, Kazaryan AM, Rosseland AR, Buanes T et al (2010) Single-centre experience of laparoscopic pancreatic surgery. Br J Surg 97:902–909. doi:10.1002/bjs.7020 3. Song KB, Kim SC, Park JB, Kim YH, Jung YS, Kim MH et al (2011) Single-center experience of laparoscopic left pancreatic
123
1606
4.
5.
6. 7.
8.
9.
10.
11.
12.
13.
14.
resection in 359 consecutive patients: changing the surgical paradigm of left pancreatic resection. Surg Endosc 25:3364–3372 Venkat R, Edil BH, Schulick RD, Lidor AO, Makary MA, Wolfgang CL (2012) Laparoscopic distal pancreatectomy is associated with significantly less overall morbidity compared to the open technique: a systematic review and meta-analysis. Ann Surg 255:1048–1059 Kooby DA, Hawkins WG, Schmidt CM, Weber SM, Bentrem DJ, Gillespie TW et al (2010) A multicenter analysis of distal pancreatectomy for adenocarcinoma: is laparoscopic resection appropriate? J Am Coll Surg 210:779–785, 786–787 Gagner M, Pomp A (1994) Laparoscopic pylorus-preserving pancreatoduodenectomy. Surg Endosc 8:408–410 Gagner M, Palermo M (2009) Laparoscopic Whipple procedure: review of the literature. J Hepatobiliary Pancreat Surg 16: 726–730 Epub 28 July 2009 Kendrick ML, Cusati D (2010) Total laparoscopic pancreaticoduodenectomy: feasibility and outcome in an early experience. Arch Surg 145:19–23 Zeh HJ, Zureikat AH, Secrest A, Dauoudi M, Bartlett D, Moser AJ (2012) Outcomes after robot-assisted pancreaticoduodenectomy for periampullary lesions. Ann Surg Oncol 19:864–870 Asbun HJ, Stauffer JA (2012) Laparoscopic vs open pancreaticoduodenectomy: overall outcomes and severity of complications using the accordion severity grading system. J Am Coll Surg 215:810–819 Sa Cunha A, Rault A, Beau C, Collet D, Masson B (2007) Laparoscopic central pancreatectomy: single-institution experience of 6 patients. Surgery 142:405–409 Rotellar F, Pardo F, Montiel C, Benito A, Regueira FM, Poveda I et al (2008) Totally laparoscopic Roux-en-Y duct-to-mucosa pancreaticojejunostomy after middle pancreatectomy: a consecutive nine-case series at a single institution. Ann Surg 247:938–944 Kang CM, Kim DH, Lee WJ, Chi HS (2011) Initial experiences using robot-assisted central pancreatectomy with pancreaticogastrostomy: a potential way to advanced laparoscopic pancreatectomy. Surg Endosc 25:1101–1106. doi:10.1007/s00464-0101324-3 Sauvanet A, Partensky C, Sastre B, Gigot JF, Fagniez PL, Tuech JJ et al (2002) Medial pancreatectomy: a multi-institutional retrospective study of 53 patients by the French Pancreas Club. Surgery 132:836–843
123
Surg Endosc (2014) 28:1601–1606 15. Christein JD, Smoot RL, Farnell MB (2006) Central pancreatectomy: a technique for the resection of pancreatic neck lesions. Arch Surg 141:293–299 16. Bassi C, Dervenis C, Butturini G, Fingerhut A, Yeo C, Izbicki J et al (2005) Postoperative pancreatic fistula: an international study group (ISGPF) definition. Surgery 138:8–13 17. Iacono C, Verlato G, Ruzzenente A, Campagnaro T, Bacchelli C, Valdegamberi A et al (2013) Systematic review of central pancreatectomy and meta-analysis of central versus distal pancreatectomy. Br J Surg 100:873–885. doi:10.1002/bjs.9136 18. Crippa S, Bassi C, Warshaw AL, Falconi M, Partelli S, Thayer SP et al (2007) Middle pancreatectomy: indications, short- and longterm operative outcomes. Ann Surg 246:69–76 19. DiNorcia J, Ahmed L, Lee MK, Reavey PL, Yakaitis EA, Lee JA et al (2010) Better preservation of endocrine function after central versus distal pancreatectomy for mid-gland lesions. Surgery 148:1247–1254. doi:10.1016/j.surg.2010.09.003 (discussion 1254–1256) 20. Adham M, Giunippero A, Hervieu V, Courbie`re M, Partensky C (2008) Central pancreatectomy: single-center experience of 50 cases. Arch Surg 143:175–180. doi:10.1001/archsurg.2007.52 (discussion 180–181) 21. Shikano T, Nakao A, Kodera Y, Yamada S, Fujii T, Sugimoto H et al (2010) Middle pancreatectomy: safety and long-term results. Surgery 147:21–29. doi:10.1016/j.surg.2009.04.036 22. Xiang GM, Tan CL, Zhang H, Ran X, Mai G, Liu XB (2012) Central pancreatectomy for benign or borderline lesions of the pancreatic neck: a single-centre experience and literature review. Hepatogastroenterology 59:1286–1289. doi:10.5754/hge11937 23. Kitasato A, Adachi T, Inokuma T, Tajima Y, Kanematsu T, Kuroki T (2012) Laparoscopic middle pancreatectomy under a pancreatic duct-navigation surgery. Hepatogastroenterology 59(120):2400 24. Kim DH, Kang CM, Lee WJ, Chi HS (2011) Robotique central pancreatectomy with pancreaticogastrostomy (transgastric approach) in a solid pseudopapillary tumor of the pancreas. Hepatogastroenterology 58:1805–1808 25. Giulianotti PC, Sbrana F, Bianco FM, Addeo P, Caravaglios G (2010) Robot-assisted laparoscopic middle pancreatectomy. J Laparoendosc Adv Surg Tech A 20:135–139
