Introduction: Pancreaticoduodenectomy (PD) is a challenging operation. Multiple studies have shown robotic PD (RPD) to be safe, with equivalent oncologic outcomes as compared to open PD (OPD). Our aim was to assess the translation of the safety profile of RPD and to determine the learning curve for RPD in a community setting. Methods: A retrospective review of 67 consecutive patients who underwent PD from January 2014 to February 2015 was performed. Results: Of the 67 PD, 55 (82%) underwent OPD and 12 (18%) RPD. Both groups were similar. There was a statistically significant difference in operative time (OT), OPD vs. RPD (OvR); (223 min vs. 337 min [P < 0.0001]), although the OT for RPD improved from 463 min (1st case) to 250 min in the last 3 cases. No statistical difference in blood loss (OvR; 435 mL vs. 335 mL [p > 0.67]), length of stay (OvR; 10.1 vs. 10.6 days [p > 0.61]) and R0 resection status (OvR; 36 vs 9 [p > 0.74]) was found. There was 1/55 (2%) mortality in the OPD group. Conclusion: Our data reflects the experience (>500 PD's) of a single high volume (>50 PD's a year) community hospital based surgeon. Similar safety profile and outcomes were achieved compared to large academic centers. For experienced HPB surgeons learning is faster and OT can be reduced to half within 10 RPD cases.

No abstract

OS rates of 71% and 53% vs. 70% and 52% at 3 and 5 years, respectively (P = 0.47) and DFS rates of 34% and 27% vs. 36% and 26% at 3 and 5 years, respectively (P = 0.64). Conclusion:In the patients who are suitable for LA, laparoscopy yields better operative outcomes without impairing long-term survival.

Management of pancreatic neuroendocrine tumors (PNET) remains controversial. We aimed to define short-and long-term outcomes associated with aggressive surgical resection. Methods: Patients with PNET who underwent definitive pancreatic resection were identified from our institutional registry (2001e2013). Demographic, pathologic, perioperative, and treatment characteristics were compared by disease extent. Outcomes included postoperative complications (ClavieneDindo grading), 90-day mortality, and actuarial disease-free and overall survival. Results: Of 123 patients, 99 (80.5%) had non-functional tumors and 5 patients (4.1%) had genetic syndromes. Eighty-six patients had localized, 19 had loco-regional, and 18 had distant metastatic disease. Median tumor size was 2.5 cm (IQR 1.5e4.5). Procedures included pancreaticoduodenectomy (44), distal pancreatectomy (55), enucleation (23), and central pancreatectomy (1). Concomitant liver-directed therapy occurred in 18 patients. Eighty patients (65%) developed complications; a majority (72.5%) of these were grade I/II. Sixteen patients (13%) developed clinically-relevant (ISGPF grade B/C) fistulas. There were no intraoperative deaths, and just one 90-day mortality related to progressive metastatic disease. When stratified by disease extent, patients with metastatic disease had higher grade tumors (45.4% vs. 10.5% vs. 3.4%, p < 0.01) and were more likely to recur (66.7% vs. 15.7% vs. 8.1%, p < 0.01) than were patients with locoregional and localized disease, respectively. At a median follow-up of 38 months (IQR 20e84), 120 (97.5%) patients were alive and 101 (82.1%) were disease-free (Figure).

No abstract