Michael Passeri, William Lyman, Amit Sastry, Allyson Cochrane, David Iannitti, Dionisios Vrochides, Erin Baker, John Martinie. Carolinas Medical Center
Background: Minimally invasive pancreaticoduodenectomy (PD) is emerging as an alternative to the traditional open technique for selected patients. The laparoscopic approach has become a viable option, but is somewhat limited by the technical complexity of the procedure, especially the reconstructive phase, which requires meticulous placement of fine sutures into delicate parenchyma and ductal structures. The robotic platform addresses many of these concerns by offering optical magnification, augmented instrument articulation, and overall greater precision with suture targeting. The aim of this study is to evaluate the postoperative outcomes of robotic PD at our institution since the procedure was introduced in 2012, and how the number of cases converted to open evolved as surgeons progressed along the learning curve.
Methods: Using the Department of HPB Surgery REDCap data repository, we compiled a list of all patients who underwent robotic pancreaticoduodenectomy since we started performing the procedure in 2012. We then dichotomized the resulting 72 patients in early and late groups. Postoperative outcomes were evaluated for those cases that were completed robotically, with emphasis on operative time, estimated blood loss, margins on malignant specimens, number of nodes excised, positive node ratio, length of stay after operation, and postoperative complication rates.
Results: A total of 72 robotic PDs were attempted. Of those, 16 (22%) were converted to open. Of the 36 in the early group, 11 (30.6%) were converted. Of the 36 in the late group, that number dropped to 5 (13.9%). Of the 56 cases that were completed robotically, 11 were performed for benign disease, while 45 were performed to excise malignant/premalignant lesions. Of those 45, R0 resection was achieved in 34 (75.6%). Median number of nodes retrieved was 18, with a positive node ratio of 0.09. Median estimated blood loss was 300cc. Median OR time was 419 minutes. Median length of stay was 8 days. Grade B or C pancreatic fistula occurred in 7.2% of cases.
Discussion: This study demonstrates that approaching PD with a robotic platform is a viable strategy in the setting of either benign or malignant disease. Conversion rates can be expected to decline as surgeons progress along the learning curve and gain familiarity with the robotic technique. The complication profile and safety metrics of cases completed robotically are in line with the standards of open PD.
Presented at the SAGES 2017 Annual Meeting in Houston, TX.
Abstract ID: 86492
Program Number: S051
Presentation Session: Robotics 1 Session
Presentation Type: Podium