Clinical Impact of Magnifying Endoscopy Combined with Narrow-Band Imaging System in Predicting Malignancy of Ampullary Tumors

Abstract

Background and Study Aims: Patients with benign ampullary tumors (ATs) can be treated by endoscopic ampullary resection, while local resection for malignant AT carries a high risk of local recurrence, even if the tumor appears confined to the papilla. There is a need for a reliable method to distinguish benign from malignant AT preoperately. Magnifying endoscopy combined with Narrow-Band Imaging (ME-NBI) can yield clear images of minute structure of the mucosa. The aim of this study was to evaluate the potential of ME-NBI in predicting malignancy of AT. Patients and Methods: Thirteen patients (mean age 66 + 9, 8 male and 5 female) with suspected ATs (size 9-25 mm) were enrolled in this study. Magnifying endoscope used in this study was GIF-Q240Z (Olympus). In 5 of 13, the ampullary changes were proven to be inflammatory by forceps biopsies and follow up of more than one year. In 4 patients, ATs were treated by endoscopic papillectomy and in 4, pancreato-duodenectomy. The study consisted of 2 phases. 1: Endoscopically or surgically resected specimens were retrieved for histopathological examination. Correlation between images of ME-NBI and histopathological findings was investigated. 2: Video-taped images of conventional duodenoscopy (CD) with dye spraying and ME-NBI were retrospectively and independently reviewed by two examiners to determine malignancy. Duodenoscopic diagnosis was made based on the presence of depression or ulceration of the mucosa. ME-NBI diagnosis was based on the findings of the phase 1 study. Accuracy was compared between CD alone, CD with biopsy and ME-NBI findings. Results: 1. ME-NBI findings were classified into the following three types: I, oval shaped villi, II, pine cone/leaf shaped villi, III, irregular/non-structured. One lesion showing type II was histopathologically diagnosed as adenoma. Seven showing type II partially with III was diagnosed as adenocarcinoma in situ or invasive adenocarcinoma. In all 5 patients with inflammatory change, ME-NBI showed type I structure. 2. Based on the phase 1 study, type III on ME-NBI was considered diagnostic of malignancy. Sensitivity/specificity of CD alone, CD with biopsy and ME-NBI were 42.9/100, 71.4/100 and 100/100%, respectively. Conclusions: In this series ME-NBI provided more accurate preoperative diagnosis of malignancy of ATs than duodenoscopic examination. It has great potential in choosing optimal option for management of AT. Background and Study Aims: Patients with benign ampullary tumors (ATs) can be treated by endoscopic ampullary resection, while local resection for malignant AT carries a high risk of local recurrence, even if the tumor appears confined to the papilla. There is a need for a reliable method to distinguish benign from malignant AT preoperately. Magnifying endoscopy combined with Narrow-Band Imaging (ME-NBI) can yield clear images of minute structure of the mucosa. The aim of this study was to evaluate the potential of ME-NBI in predicting malignancy of AT. Patients and Methods: Thirteen patients (mean age 66 + 9, 8 male and 5 female) with suspected ATs (size 9-25 mm) were enrolled in this study. Magnifying endoscope used in this study was GIF-Q240Z (Olympus). In 5 of 13, the ampullary changes were proven to be inflammatory by forceps biopsies and follow up of more than one year. In 4 patients, ATs were treated by endoscopic papillectomy and in 4, pancreato-duodenectomy. The study consisted of 2 phases. 1: Endoscopically or surgically resected specimens were retrieved for histopathological examination. Correlation between images of ME-NBI and histopathological findings was investigated. 2: Video-taped images of conventional duodenoscopy (CD) with dye spraying and ME-NBI were retrospectively and independently reviewed by two examiners to determine malignancy. Duodenoscopic diagnosis was made based on the presence of depression or ulceration of the mucosa. ME-NBI diagnosis was based on the findings of the phase 1 study. Accuracy was compared between CD alone, CD with biopsy and ME-NBI findings. Results: 1. ME-NBI findings were classified into the following three types: I, oval shaped villi, II, pine cone/leaf shaped villi, III, irregular/non-structured. One lesion showing type II was histopathologically diagnosed as adenoma. Seven showing type II partially with III was diagnosed as adenocarcinoma in situ or invasive adenocarcinoma. In all 5 patients with inflammatory change, ME-NBI showed type I structure. 2. Based on the phase 1 study, type III on ME-NBI was considered diagnostic of malignancy. Sensitivity/specificity of CD alone, CD with biopsy and ME-NBI were 42.9/100, 71.4/100 and 100/100%, respectively. Conclusions: In this series ME-NBI provided more accurate preoperative diagnosis of malignancy of ATs than duodenoscopic examination. It has great potential in choosing optimal option for management of AT.