Influence of margin histology on development of pancreatic fistula following pancreatoduodenectomy

Abstract Background Postoperative pancreatic fistula (POPF) is a potentially debilitating complication following pancreatoduodenectomy (PD). There are limited data correlating pancreatic parenchymal histopathologic features specifically fat and fibrosis content with development of POPF after PD. Methods Patients who underwent PD (January 2010-May 2015) with archived pathologic slides were included. Each pancreatic neck transection margin was histologically graded for fat and fibrosis, scored from 0 to 4, and grader was blinded to clinical outcomes. Main pancreatic duct diameter and duct wall thickness were microscopically measured. Patients were dichotomized into high and low categories with respect to pancreatic fat and fibrosis and primary outcome of POPF. Results Of 301 patients, 24 developed POPF (8.0%). One hundred ten patients (36.5%) had low fat (score versus 3.3%, P  = 0.005). Low fibrosis (odds ratio [OR] 4.29, 95% confidence interval [CI] 1.56-11.7, P  = 0.005), nonpancreatic adenocarcinoma pathology (OR 3.25, 95% CI 1.25-8.43, P  = 0.02), and increased body mass index (BMI) (OR 1.11, 95% CI 1.03-1.12, P  = 0.007) were associated with POPF development on univariate analysis. Low fibrosis and increased BMI remained independently associated on multivariate analysis. High fat content was frequently concurrently identified in specimens with high fibrosis (67.8%). Surgeon-described gland consistency did not correlate with histopathologic findings (Spearman's rank correlation coefficients of −0.144 and 0.304, respectively) or to incidence of POPF. No patient who underwent preoperative chemotherapy developed POPF ( n  = 30, 10%). Conclusions Low pancreatic neck fibrosis content and increased patient BMI are associated with increased rates of POPF following PD, while pancreatic fat content does not appear to influence this outcome. Pancreatic neck fat and fibrosis often coexist in the same specimen. The association between preoperative chemotherapy and low POPF rates needs further examination. Frozen section analysis of pancreatic neck margin for fibrosis content may be more accurate than surgeon assessment in identifying patients at risk for POPF. These assessments can potentially guide therapeutic interventions, including selective prophylactic drain placement and use of postoperative somatostatin analog therapy.