Robotic pylorus-preserving pancreaticoduodenectomy: Technical considerations

Case Report: Robotic pylorus-preserving pancreatoduodenectomy for periampullary rhabdomyosarcoma in a 3-year-old patient

Periampullary neoplasm is rare in pediatric patients and has constituted a strict indication for pancreatoduodenectomy (PD), which is a procedure sporadically reported in the literature among children. Robotic PD has been routinely performed for periampullary neoplasm in periampullary neoplasm, but only a few cases in pediatric patients have been reported. Here, we report the case of a 3-year-old patient with periampullary rhabdomyosarcoma treated with robotic pylorus-preserving PD and share our experience with this procedure in pediatric patients. A 3-year-old patient presented with obstructive jaundice and a mass in the pancreatic head revealed by imaging. A laparoscopic biopsy was performed. Jaundice progressed with abdominal pain and elevated alpha-amylase leading to urgent robotic exploration in which a periampullary neoplasm was revealed and pathologically diagnosed as rhabdomyosarcoma by frozen section examination. After pylorus-preserving PD, we performed a conventional jejunal loop following a child reconstruction, including an end-to-end pancreaticojejunostomy, followed by end-to-side hepaticojejunostomy and duodenojejunostomy. Delayed gastric emptying (DGE) presented with increasing drain from the nasogastric tube (NGT) a week after the surgery and improved spontaneously within 10 days. In a 13-month follow-up until the present, our case patient recovered well without potentially fatal complications, such as pancreatic fistula. Robotic PD in pediatric patients was safe and effective without intra- or postoperative complications.

Surgical and oncological outcomes of our first 59 cases of robotic pancreaticoduodenectomy

PURPOSE

Robotics has shown encouraging results for a number of technically demanding abdominal surgeries including pancreaticoduodenectomy, which has originally represented a relative contraindication to the application of the minimally-invasive technique. We aimed to investigate the perioperative, clinicopathologic, and oncological outcomes of robot-assisted pancreaticoduodenectomy by assessing a consecutive series of totally robotic procedures.

METHODS

All consecutive patients who underwent robotic pancreaticoduodenectomy were included in the present analysis. Perioperative, clinicopathologic and oncological outcomes were examined. In order to investigate the role of the learning curve, surgical outcomes were also used to compare the early and the late phase of our experience.

RESULTS

A total of 59 patients underwent surgery. Median hospital stay was 9 days (5 - 110), with an overall morbidity and mortality of 37% and 3%, respectively. Of note, the rate of clinically relevant pancreatic fistula was 11.8%. R0 resections were achieved in 96% of patients and the 3-year disease-free and overall survivals were 37.2 and 61.9%, respectively. Overall, surgical outcomes did not vary significantly between the first and the late phase of the series.

CONCLUSIONS

Robotic pancreaticoduodenectomy can be performed competently. It satisfies all features of oncological adequacy and may offer a number of advantages over standard procedures in terms of surgical results.

Technical considerations for the fully robotic pancreaticoduodenectomy

Minimally invasive surgery, including robotic surgery, has become the standard of care for many abdominal procedures. However, the technical complexity associated with pancreaticoduodenectomy (PD) due to the anatomic location and oncologic characteristics of pancreatic tumors has hindered the widespread application of minimally invasive techniques to this procedure. Recent studies have reported that for experienced surgeons, the application of robotic techniques to PD is associated with equivalent oncologic outcomes and rates of complication when compared to an open operation, and may be associated with accelerated surgical recovery. Despite these encouraging results, robotic PD (RPD) is a procedure attempted by a small group of pancreatic surgeons, leading to the great heterogeneity in the techniques used to perform this operation. Herein we describe our technique for fully RPD and demonstrate its execution with a video supplement.

Robotic pancreatoduodenectomy with vascular resection

PURPOSE

This study aims to define the current status of robotic pancreatoduodenectomy (RPD) with resection and reconstruction of the superior mesenteric/portal vein (RPD-SMV/PV).

METHODS

Our experience on RPD, including RPD-SMV/PV, is presented along with a description of the surgical technique and a systematic review of the literature on RPD-SMV/PV.

RESULTS

We have performed 116 RPD and 14 RPD-SMV/PV. Seven additional cases of RPD-SMV/PV were identified in the literature. In our experience, RPD and RPD-SMV/PV were similar in all baseline variables, but lower mean body mass and higher prevalence of pancreatic cancer in RPD-SMV/PV. Regarding the type of vein resection, there were one type 2 (7.1 %), five type 3 (35.7 %) and eight type 4 (57.2 %) resections. As compared to RPD, RPD-SMV/PV required longer operative time, had higher median estimated blood loss, and blood transfusions were required more frequently. Incidence and severity of post-operative complications were not increased in RPD-SMV/PV, but post-pancreatectomy hemorrhage occurred more frequently after this procedure. In pancreatic cancer, RPD-SMV/PV was associated with a higher mean number of examined lymph nodes (60.0 ± 13.9 vs 44.6 ± 11.0; p = 0.02) and with the same rate of microscopic margin positivity (25.0 % vs 26.1 %). Mean length or resected vein was 23.1 ± 8.08 mm. Actual tumour infiltration was discovered in ten patients (71.4 %), reaching the adventitia in four patients (40.0 %), the media in two patients (20.0 %), and the intima in four patients (40.0 %). Literature review identified seven additional cases, all reported to have successful outcome.

CONCLUSIONS

RPD-SMV/PV is feasible in carefully selected patients. The generalization of these results remains to be demonstrated.