Diagnostic accuracy of endoscopic ultrasonography-guided tissue acquisition prior to resection of pancreatic carcinoma: a nationwide analysis

Abstract

Introduction: Endoscopic ultrasonography guided tissue acquisition (EUS + TA) is used to provide a tissue diagnosis in patients with suspected pancreatic cancer. Key performance indicators (KPI) for these procedures are rate of adequate sample (RAS) and sensitivity for malignancy (SFM). Aim: assess practice variation regarding KPI of EUS + TA prior to resection of pancreatic carcinoma in the Netherlands. Patients and methods: Results of all EUS + TA prior to resection of pancreatic carcinoma from 2014–2018, were extracted from the national Dutch Pathology Registry (PALGA). Pathology reports were classified as: insufficient for analysis (b1), benign (b2), atypia (b3), neoplastic other (b4), suspected malignant (b5), and malignant (b6). RAS was defined as the proportion of EUS procedures yielding specimen sufficient for analysis. SFM was calculated using a strict definition (malignant only, SFM-b6), and a broader definition (SFM-b5+6). Results: 691 out of 1638 resected patients (42%) underwent preoperative EUS + TA. RAS was 95% (range 89–100%), SFM-b6 was 44% (20–77%), and SFM-b5+6 was 65% (53–90%). All centers met the performance target RAS>85%. Only 9 out of 17 met the performance target SFM-b5+6 > 85%. Conclusion: This nationwide study detected significant practice variation regarding KPI of EUS + TA procedures prior to surgical resection of pancreatic carcinoma. Therefore, quality improvement of EUS + TA is indicated.