Performance Of Eus-Guided Fna Of Solid Pancreatic Lesions According To The Level Of Operator-Experience

Aims The amount of training before the independent performance of EUS-FNA of solid pancreatic lesions has been poorly described. In previous studies, rapid on-site cytopathological examination (ROSE) to evaluate the FNA sample adequacy was used. We aimed to evaluate the number of interventions needed to reach acceptable competence in EUS-guided FNA of solid pancreatic lesions in centers without ROSE.

Methods Patients with solid pancreatic masses who underwent EUS-FNA between 07/2016-09/2020 in our tertiary center were identified in a prospectively collected database. Examinations were performed by three operators without prior experience in EUS tissue sampling. The first 15-20 FNAs were performed under the direct supervision of an experienced endosonographer. According to the number of EUS-FNAs performed, the following levels of experience were defined (group A-F): ≤10,11-20,21-30,31-40,41-50,>50 interventions. Complications were assessed according to the ESGE guidelines. The final diagnosis was determined by cyto-histopathology or clinical follow-up with a combination of tumor-markers and radiological findings.

Results 150 FNAs in 137 patients (mean age 67 years, 63.3 % males) with solid pancreatic lesions were analyzed: 76.7 % malignant (60.7 % adenocarcinomas), and 23.3 % benign masses. Similar baseline characteristics were found for all patient groups examined by levels of operator-experience. The diagnostic performance of EUS-FNA improved rapidly after performing at least 40 interventions (Table). Best accuracy and sensitivity for malignancy were observed in experts (group F) and were higher especially compared to endosonographers with less supervision (group C-D): 81.8 % vs. 55 % (p = 0.02), and 78.3 % vs. 58.1 % (p = 0.18), respectively. Severe complications were rare (2.7 %), and 3/4 of them occurred early in the training process (group A-B).

Experience (No. EUS-FNAs)	Accuracy	Sensitivity for malignancy	Complication rate
(A) ≤ 10 (B) 11-20 (C) 21-30 (D) 31-40 (E) 41-50 (F) > 50	(A) 60 % (18/30) (B) 77.8 % (21/27) (C) 55 % (11/20) (D) 55 % (11/20) (E) 75 % (15/20) (F) 81.8 % (27/33)	(A) 64 % (16/25) (B) 76.5 % (13/17) (C) 53.8 % (7/13) (D) 61.1 % (11/18) (E) 76.5 % (13/17) (F) 78.3 % (18/26)	(A) 3.3 % (1/30) (B) 7.4 % (2/27) (C) 0 % (0/20) (D) 0 % (0/20) (E) 0 % (0/20) (F) 2.8 % (1/36)

Conclusions Despite the lack of real-time evaluation of sample adequacy, the accuracy and sensitivity for malignancy improved after performing at least 40 EUS-guided FNAs in solid pancreatic lesions. Early in the learning process (< 20 EUS-FNA), additional attention to avoid complications is warranted.

Citation Razpotnik M, Bota S, Essler G et al. eP407 PERFORMANCE OF EUS-GUIDED FNA OF SOLID PANCREATIC LESIONS ACCORDING TO THE LEVEL OF OPERATOR-EXPERIENCE. Endoscopy 2021; 53: S230.

Publication History

Article published online:
19 March 2021