Indications and Perioperative Outcomes for Pancreatectomy with Arterial Resection

Pancreatectomy with arterial resection (AR) is performed infrequently. As indications evolve, we evaluated indications, outcomes, and predictors of mortality, morbidity, and survival after AR.

We performed a single-institution review of elective pancreatectomies with AR (from July1990 to July 2017). Univariate and multivariate analyses were performed for predictors of outcomes and survival.

A total of 111 patients underwent pancreatectomy with AR including any hepatic (54%), any celiac (44%), any superior mesenteric (14%), or multiple ARs (14%), with revascularization in 55%. The majority of cases were planned (77%) and performed post-2010 (78%). Overall 90-day major morbidity (≥grade III) and mortality were 54% and 13%, respectively, due to post-pancreatectomy hemorrhage (PPH), postoperative pancreatic fistula (POPF), or ischemia in the majority of cases. There was a significant decrease in mortality post-2010 (9% vs 29%, p = 0.02), and this was protective on multivariate analysis (odds ratio [OR] 0.1, p = 0.004); PPH increased mortality (OR 6.1, p < 0.001). Post-pancreatectomy hemorrhage was associated with major morbidity (OR 5.1, p = 0.005), reoperation (OR = 23.0, p = 0.004), ICU (OR 5.5, p < 0.001), and readmission (OR 2.6, p = 0.004). Other morbidity predictors were AR with graft (OR 4.0, p = 0.031) and POPF (OR 3.1, p = 0.003). Median survival was 28.5 months and improved for ductal adenocarcinoma after neoadjuvant chemotherapy (p = 0.038). There were no differences in survival based on AR type.

Regardless of indication or type, pancreatectomy with AR is associated with risks greater than standard resections. Mortality has decreased in the modern era; however, morbidity remains high from hemorrhagic, fistula, or ischemia-related complications. Mitigation measures are needed if advanced resections are considered with increasing frequency given the potential oncologic benefit of AR in selected cases after modern chemotherapy.