Objectives: Studies correlating endoscopic ultrasound (EUS) with histopathology for chronic pancreatitis (CP) are limited by small sample size, and/or inclusion of many patients without CP, limiting applicability to patients with painful CP. The aim of this study was to assess correlation of standard EUS features for CP with surgical histopathology in a large cohort of patients with non-calcific CP (NCCP).
Methods: Adult patients undergoing total pancreatectomy and islet autotransplantation (TPIAT) for NCCP, between 2008 and 2013, with EUS <1 year before surgery. Histology from resected pancreas at the time of TPIAT (from head, body, and tail) was graded by a GI pathologist blinded to the EUS features. A fibrosis score (FS) ≥2 was abnormal, and FS≥6 was considered severe fibrosis. A multivariate regression analysis for the EUS features predicting fibrosis, after taking age, sex, smoking, and body mass index into consideration, was performed. A quantitative receiver operating characteristic (ROC) curve analysis was performed and Spearman rank correlation co-efficient (r) was calculated.
Results: 68 patients (56 females, mean±s.d. age-38.77±10.92) underwent TPIAT for NCCP with pre-operative EUS. ROC curve showed that four or more EUS features provided the best balance of sensitivity (61%), specificity (75%), and accuracy (63%). Although significant, correlation between standard EUS features and degree of fibrosis was poor (r=0.24, P<0.05). Multivariate regression analysis showed that main pancreatic duct irregularity was the only independent EUS feature (P=0.02) which predicted CP.
Conclusions: Correlation between standard EUS features and histopathology is poor in NCCP. MPD irregularity is an independent predictor for NCCP.