Emanuele Federico Kauffmann

The Learning Curve in Robotic Pancreaticoduodenectomy

Background/Purpose: Few data are available on the learning curve (LC) in robot-assisted pancreaticoduodenectomy (RAPD) and no study specifically addresses the LC of a single surgeon. Methods: The LC of a single surgeon in RAPD was determined using the cumulative sum method, based on operative time (OT). Data were extracted from a prospectively maintained database and analyzed retrospectively considering all events occurring within 90 days of index operation. Results: Seventy RAPD were analyzed. One operation was converted to open surgery (1.4%). One patient died within 30 days (1.4%) and one within 90 days (2.8%). Postoperative complications occurred in 53 patients (75.7%) and exceeded Clavien-Dindo grade IIIb in 7 patients (10%). OT dropped after 33 operations from a mean of 564 ± 101.7 min to a mean of 484.1 ± 77.9 min (p = 0.0005) and was associated to reduced incidence of delayed gastric emptying (72.7 vs. 48.7%; p = 0.039). The rate of hospital readmission improved after 40 operations from 20.0 (8 of 40) to 3.3% (1 of 30) (p = 0.04). Conclusions: RAPD was safely feasible in selected patients. OT dropped after the first 33 operations and was associated with reduced rate of delayed gastric emptying. Readmission rate improved after 40 operations.

Total Robotic Pancreaticoduodenectomy

Context The “da Vinci” surgical system reintroduces much of the operative dexterity lost during laparoscopic operations and offers the unique opportunity to verify if pancreaticoduodenectomy (PD) can be safely performed through a minimally invasive approach. Objective We report our technique for total robotic PD employed in 39 consecutive patients. This experience was earned at a high-volume center of pancreatic surgery, having extensive experience in advanced laparoscopy and robotic surgery. Methods Our technique for total robotic PD is unique in several respects: pure laparoscopy is not used at any stage; only the right colonic flexure is mobilized; a total of five ports are used; the camera port is placed along the right pararectal line to allow optimal view of the uncinate process (UP); the third robotic arm is placed on the patient’s left side and it is used to “hang” the duodenum during dissection of the UP; the gallbladder is used to retract the liver; the first jejunal loop is fully mobilized but it is not sectioned until the specimen is ready for removal, to facilitate jejunal rotation behind the mesenteric vessels. Results No PD was converted to open surgery or laparoscopy, despite 3 patients required segmental resection of the mesenteric vein and reconstruction by a jump graft. Mean operative time was 597 minutes (range: 420-960 minutes). Thirty-day operative mortality was nil. No pseudoaneurysm of the gastroduodenal artery was noted. Only 4 patients developed grade B pancreatic fistulas and none grade C fistulas. Mean hospital-stay was 23 days (range: 10-86 days). Malignant tumors were diagnosed in 51% of the patients. Overall, the mean number of lymph nodes retrieved was 32 (range 15-76). None of the margins was positive. Conclusions In selected patients total robotic PD is feasible. As compared to hybrid techniques, coupling laparoscopic dissection with robotic reconstruction, a total robotic procedure spares unnecessary dissections and allows optimal control of large peripancreatic vessels permitting segmental vein resection and tailored reconstruction. Technology refinements and improvement of surgical technique could make robotic PD an appealing alternative to open PD in selected patients.

Analysis of Instrument Traffic During Laparoscopic Robot-Assisted Pancreaticoduodenectomy

Context Robotic surgery entails specific issues that are not present, or are not equally relevant, in open surgery or conventional laparoscopy. Instrument traffic (IT) is one of such issues. IT is the time during which surgery is paused because the surgeon at the console is waiting for the action of the surgeon the table (e.g., instrument change, camera cleaning, introduction/withdrawal of needles). Objective We provide the first objective evaluation of IT during robotic pancreaticoduodenectomy (PD). Methods The operative videos of 12 robot-assisted PDs were reviewed to define IT. The analysis included: crude IT time (CITT), relative IT time (RITT) (defined as the percentage of operative time spent for IT), number of robotic instruments changes (RIC), time spent for RIC (TRIC), number of pure laparoscopic actions (PLA), and time spent for PLA (TPLA). Figures were estimated for the entire operation as well as for dissection and reconstruction phases. Details on pancreaticojejunostomy (PJ) or hepaticojejunostomy (HJ) were related to IT to define their relative impact on operative time. Results Mean operative time was 517 min (range 420-600 min). Mean CITT was 3,681.6 sec (RITT 11.89%). Mean RIC or PLA was 315.7. Each RIC or PLA paused surgery for 11.8 sec. Mean RIC was 184.4 (TRIC 2,633.8 sec). Mean PLA was 131.4 (TPLA 1,039.5 sec). Mean dissection time was 326.9 min. Mean CITT was 2,095.1 sec (RITT 10.68%). Mean RIC was 105.8 (TRIC 1,645.2 sec). Mean PLA was 35.7 (TPLA 382.5 sec). Each RIC or PLA paused surgery for 14.4 sec. PJ was made by invaginating technique or duct-to-mucosa. The last one required fewer stitches, but did not reduced CITT or RIC. HJ was performed using either 4 half running sutures or interrupted external stitches plus inner half running sutures. Despite similar CITT the former technique was associated with fewer RIC. Conclusions Some 12% of operative time of laparoscopic robot-assisted PD is wasted because of IT. Since in this series operative time of robotic PD averaged 517 minutes, IT prolonged surgery of more than one hour. Technology improvements and/or refinements in surgical technique are expected to reduce IT during robotic PD.

Incidence and reasons of pancreatic resection in patients with asymptomatic serous cystadenoma

BACKGROUND/OBJECTIVES Despite diagnostic refinements, pancreatic resection (PR) is eventually performed in some patients with asymptomatic serous cystadenoma (A-SCA). The aim of this study was to define incidence and reasons of PR in A-SCA. METHODS A retrospective analysis of a prospectively maintained database was performed for all the patients referred for pancreatic cystic lesions (PCL) between January 2005 and March 2016. RESULTS Overall, there were 1488 patients with PCL, including 1271 (85.4%) with incidental PCL (I-PCL). During the study period referral of I-PCL increased 8.5-fold. Surgery was immediately advised in 94 I-PCL (7.3%) and became necessary later on in 11 additional patients (0.9%), because of the development of symptoms. Overall, PR was performed in 105/1271 patients presenting with I-PCL (8.2%), including 27 with A-SCA (2.1%). All patients with A-SCA underwent ultrasonography and contrast-enhanced computed tomography. Magnetic resonance imaging was performed in 21 patients (77.8%), 18 F-FDG positron emission tomography in 8 (29.6%), endoscopic ultrasonography (EUS) in 2 (7.4%), and EUS-guided fine needle aspiration (EUS-FNA) in 1 (3.7%). These studies demonstrated a combination of atypical features such as solid tumor (3; 11.1%), oligo-/macrocystic tumor (24; 88.8%), mural nodules (14; 51.8%), enhancing cyst walls (17; 62.9%), dilation of the main pancreatic duct (3; 11.1%), and upstream pancreatic atrophy (1; 3.7%). Additionally, 14/27 patients (51.8%) were females with oligo-/macrocystic tumors located in the body-tail of the pancreas. CONCLUSIONS Management of patients with A-SCA entails a small risk of PR especially when these tumors demonstrate atypical radiologic features associated with confounding anatomic and demographic characteristics.

Pancreatic Metastasis from Colorectal Cancer

Context Pancreatic metastases are rare (2% of all pancreatic carcinomas). Very few cases about surgical treatment of colorectal cancer metastases to the pancreas are reported. Case report We report a case of single colorectal cancer metastasis to the pancreas managed by distal splenopancreatectomy in a patient undergone to left hemicolectomy for the primary tumor eight years before and to middle lung lobectomy for metastasis one year before. A 61-year-old asymptomatic woman with a history of colorectal cancer was admitted to our department after that during the oncological imaging follow-up a thoracic-abdominal contrast-enhanced computed tomography (CT) demonstrated a single 25 mm hypodense lesion in the pancreatic tail. She also presented high levels of CEA (61.7 ng/mL) and CA 19-9 (82.9 U/mL) before the admission. Eight years before the patient underwent to left hemicolectomy for a B2-Dukes classification colorectal cancer. The resected margins were free of tumor and no regional lymph nodes were positive. One year before the patient underwent to a lung lobectomy for a single 30 mm pulmonary metastases. Considering history and imaging findings, the pancreatic lesion was suspected a colorectal cancer metastasis. A distal splenopancreatectomy was performed. The patient was discharged in healthy conditions. Final pathology disclosed the pancreatic lesion was a colorectal cancer metastasis (CD20+, CK7-) with infiltration of the peri-pancreatic adipose tissue. The resected margins were free of tumor and no lymph nodes were metastasized. The patient is still alive. Conclusion Metastases to the pancreas are commonly considered rare, especially those from colorectal cancer. The improvement of imaging techniques has led to an increase of diagnoses and surgical procedures for metastases to the pancreas. Secondary tumors may be considered in the differential diagnosis of primitive pancreatic lesions. The diagnosis may be facilitated by clinical history and serum markers assessment. Metastatic colorectal cancer to the pancreas is an indication for pancreatic resection to increase the overall survival and, as palliative procedure, to treat symptoms like jaundice and pain.

One-Hundred and Six Robot-Assisted Pancreatectomies

Context Laparoscopy has revolutionized abdominal surgery becoming the standard approach for many operations. The “da Vinci” surgical system overcomes most of the inherent technical limitations of laparoscopy. Objective We test whether the robotic approach can improve the outcome of pancreatic resections, which often require challenging dissection and complex digestive reconstructions. Methods One-hundred and six consecutive robotic pancreatic resections were performed between October 2008 and June 2012. There were 40 males and 66 females (62%), with a mean age of 57 years (range 21-80 years) and a mean BMI of 24.6 Kg/m 2 . Thirty-nine patients underwent pancreaticoduodenectomy (PD) (37%), 47 distal pancreatectomy (DP) (44%), 10 total pancreatectomy (10%), 7 tumor enucleation (6%) and 3 central pancreatectomy (3%). Since our activity spans over about a 4-year period, data were analyzed according to the time of surgery, to verify progress in the learning curve: 17 patients were operated on between October 2008 and September 2009, 22 patients between October 2009 and September 2010, 32 patients between October 2010 and September 2011 and 35 patients during the last 9 months (from October 2011 to June 2012). Results No patient was converted to laparoscopy or open surgery. Mean operative time (OT) was 442.8 minutes. In the first period OT was 512 min for PD and 420 for DP. The mean number of lymph nodes examined (LN) was 16.8; 31.2 for PD and 11.9 for DP. Pancreatic fistula (PF) occurred in 41% of the patients. In the second, OT was 596 min for PD and 402 for DP. The LN was 16.7; 27.2 for PD and 10.0 for DP. PF was amounted 36.3%. In the third, OT was 583 min for PD and 288 for DP. The LN was 28.7; 36.0 for PD and 19.1 for DP. PF was amounted 36.6%. In the fourth, OT was 590 min for PD and 250 for DP. The LN was 30; 32 for PD and 20 for DP. PF was amounted 35%. Fifty-six benign/low-grade tumors and 50 cancers were diagnosed. Surgical margins were all negative. Post-operative mortality was nil, morbidity was 56% and mean hospital stay was 16 days. Conclusions Robot-assisted pancreatic resections can be safely performed in selected patients. Despite the existence of a learning curve, experienced pancreatic surgeons are not expected to pay to robotics the same price that they would have been asked for by laparoscopy.