Decreased national pancreatic fistula rates with minimally invasive pancreaticoduodenectomy: An evolving benefit over the open approach?

Abstract

Presenter: Jorge Zarate Rodriguez MD | Washington University, St. Louis Background: Pancreatic fistula is the major driver of morbidity and mortality after pancreaticoduodenectomy (PD), and there is a steep learning curve associated with minimally invasive pancreaticoduodenectomy (MIPD). We hypothesized that national MIPD outcomes would improve over time, and would demonstrate a benefit over open pancreaticoduodenectomy (OPD) with increasing experience. Methods: All patients undergoing elective MIPD (laparoscopic, robotic, and laparoscopic/robotic converted to open) or OPD and accrued into the pancreatectomy targeted NSQIP database between 2014 and 2017 were included in the study. Patient variables and outcomes after MIPD versus OPD were compared. Clinically relevant postoperative pancreatic fistula (CR-PF) rates were compared between MIPD and OPD for each individual year in the database. Patients were stratified into era 1 (2014-2015) and era 2 (2016-2017). MIPD patients were propensity score matched to OPD patients (1:3) and their outcomes compared within each era. Results: 13,373 patients underwent PD between 2014-2017: 12,303 (92.0%) OPD and 1,070 MIPD (8.0%). Patients undergoing MIPD were less likely to be jaundiced (35.5% vs 43.7%, p<0.001), have recent weight loss (9.8% vs 15.4%, p<0.001), have T4 tumors (53.8% vs 57.7%, p<0.001), undergo neoadjuvant radiation therapy (6.3% vs 8.1%, p=0.042) and undergo vascular resection (13.6% vs 17.7%, p<0.001), but were more likely to be white (81.8% vs 7.7%, p<0.001), have a drain placed at surgery (96.1% vs 87.5%, p<0.001), and have soft pancreas gland texture (40.7% vs 35.0%, p=0.003). When comparing CR-PF rates after MIPD to OPD in each individual year, CR-PF rates went from being significantly higher after MIPD compared to OPD in 2014 (18.7% vs 13.5%, p= 0.050) to being significantly lower than OPD in 2017 (8.6% vs 13.8%, p=0.009) (Figure). After 1:3 propensity score matching patients undergoing MIPD (n=440) to those undergoing OPD (n=1320) in era 1 (years 2014-2015), MIPD was associated with a significantly increased rate of 30-day readmission (20.5% vs 14.8%, p=0.005) with no other differences in postoperative outcomes. After 1:3 propensity score matching patients undergoing MIPD (n=630) to those undergoing OPD (n=1890) in era 2 (years 2016-2017), MIPD was associated with significantly decreased rates of CR-PF (9.8% vs 14.8%, p=0.002), organ space infection (12.4% vs 16.9%, p=0.007), systemic sepsis (5.7% vs 8.9%, p=0.011), and myocardial infarction (0.2% vs 1.0%, p=0.046) along with a significantly decreased length of postoperative stay (9.7 days vs 10.7 days, p=0.013). Conclusion: National MIPD outcomes have significantly improved over time in the NSQIP pancreatectomy database, and MIPD has significant perioperative benefits over OPD in recent years.