Correlation Between EUS and Histopathology in 50 Patients Undergoing Total Pancreatectomy (TP) with Islet Autotransplantion (IAT) for Minimal Change Chronic Pancreatitis (MCCP): ACG Auxiliary Award

Abstract

Purpose: Diagnosis of minimal change (non-calcific) chronic pancreatitis (MCCP) is challenging and controversial. Endoscopic ultrasound (EUS) has been widely used for diagnosing MCCP using 9 standard criteria, however, these criteria are derived from correlation between EUS and histology from patients (pts) undergoing resection for conditions other than MCCP. The aim of our study was to investigate the correlation between EUS findings and histopathology in pts undergoing total pancreatectomy with islet autotransplantation (TP-IAT) for painful MCCP. Methods: We retrospectively compared EUS findings (obtained <1 year prior to surgery) with histopathology of the pancreas in 50 adults and children with MCCP (no calcifications on abdominal CT) from a total of 141 pts undergoing TP-IAT at a single center from 1/2008 through 7/2010. EUS abnormalities were classified using 9 standard criteria. Pts were selected for TP-IAT by a multidisciplinary committee based on intractable disabling pancreatic pain, absence of correctable etiology, refractory to maximal endoscopic therapy, plus MCCP based on the following criteria: documented history of acute recurrent pancreatitis, or abnormality in at least 2 of the following 3 tests: 1) EUS (≥4/9 criteria), 2) secretin stimulated MRCP with ductographic abnormalities or decreased T1 parenchymal signal, 3) abnormal endoscopic pancreatic function test (peak bicarbonate concentration <80 mM/L). Histology was obtained from resected pancreas by wedge biopsy of head, body and tail when possible (mean # of biopsies 2.6). Histopathology was evaluated by a blinded pancreas pathologist. MCCP was determined as present histologically when fibrosis, atrophy or inflammation was found in at least one post resection biopsy. Results: MCCP using composite of any of 3 abnormalities was found at histopathology in 45/50 (90%) pts. All 45 (100%) pts had fibrosis. The 5 pts (10%) with normal histology all had a documented history of acute recurrent pancreatitis in addition to disabling pain. Of pts with histologically confirmed MCCP, 27/45 (60%) had ≥4/9 criteria on EUS. 18/21 (85%) pts with ≤3 EUS criteria and 4/5 (80%) pts with 0/9 EUS criteria had both fibrosis and MCCP by histopathology. 2/29 (7%) pts with ≥4/9 EUS criteria had normal histology. AUC was 0.593, Kappa Coefficients <0.11 for all cutoff points. Negative predictive value of a normal EUS was only 38%, and positive predictive value of abnormal EUS only exceeded 72% at ≥6 criteria. Conclusion: Correlation between EUS and histopathology of minimal change chronic pancreatitis is poor in pts undergoing TP-IAT. Normal or nearly normal EUS cannot be used to exclude MCCP, and abnormal EUS alone is probably not sufficient for the diagnosis unless ≥6 criteria are present.