Lessons learned from the introduction of laparoscopic pancreticoduodenectomy

Introduction: Laparoscopic pancreaticoduodenectomy (LPD) is rapidly gaining popularity. Open PD has been described a technically demanding operation. In the Netherlands, pancreatic resections are restricted to centres with volume 20 PDs/year. In 2016 we started with the LPD. In this study we aimed to identify the pitfalls of the introduction of minimal-invasive PD. Methods: All operations were done by two senior surgeons. All surgeons had extensive laparoscopic experience with each more than 500 advanced laparoscopic procedures (colorectal, liver, spleen, upper GI, nephrectomy). The first 5 laparoscopic PDs were supervised by a proctor. We retrospectively reviewed the first 20 laparoscopic PD cases. All data were obtained from our prospectively maintained database. The primary outcome was all anastomosis-related morbidity. Results: After 2 complicated cases and discussion with an international expert, we decided to change from a full laparoscopic approach to a hybrid approach in which the anastomoses were done through a small midline laparotomy. In 11 patients a full laparoscopic approach was attempted, 4 patients were converted. In 7 fully laparoscopic procedures, 6 patients experienced anastomosis-related complications. There was one in-hospital mortality in this group. In the 9 patients a hybrid technique was used. There was no anastomosis-related morbidity in the hybrid group (p = 0.01, Fischer's exact test). There were no conversions en there was no mortality in the hybrid group. Conclusions: Anastomosis- related complications were frequently seen after the introduction of the full laparoscopic pancreaticoduodenectomy. The hybrid approach may be safer in the earliest phase of the learning curve of minimal-invasive pancreaticoduodenectomy.