On Site Cytopatologist Assessment For Endosonography Guided Biopsy: Always, Never, On Demand?

Background: Endoscopic ultrasound guided fine needle aspiration (EUS-FNA) is a modern technology that allows one to obtain a tissue diagnosis confirmation in peri-intestinal lesions, with a mimimum risk of complications. On site cytology assessment has been shown to improve EUS-FNA diagnostic accuracy by 10-15% (when compared with no on site assessment). However, it remains unclear if this diagnostic benefit applies to all type of lesions biopsied (solid vs cystic, adenopathy vs pancreas mass). Aim: To prospectively compare the diagnostic accuracy of EUS-FNA with and without on site cytology assessment in different type of lesions. Material and Methods: Period of inclusion: 11/03-11/05. Patients (PT) with a lesion amenable for EUS-FNA and no prior tissue diagnosis were included. EUS exam (conscious sedation): GF-UCT160-OL5 Olympus/22 Gauge needle. Diagnostic accuracy of EUS-FNA was prospectively evaluated in 2 cohorts of PT seeking care at Hospital A (on site EUS-FNA cytology assessment available/by cytopathologist) and Hospital B (on site EUS-FNA cytology assessment not available), both evaluated by a single endosonographer (EVS). Lesions sampled (EUS-FNA) were categorized according to type of lesion and location (1. lympadenopathy, 2. pancreas solid mass and 3. pancreas cystic mass) for analysis (EUS-FNA of the first 50 PT in each category in Cohort A and in Cohort B, durign the study period). Statistical analysis: Sensitivity, specificity and accuracy of EUS-FNA in Cohort A and B were compared (Fisher Test). Number of passes performed were analyzed by means of ANOVA and student t Test. Level of significance: 0.05. Results: Patient age, sex, indication for EUS-FNA, and history of chronic pancreatitis did not differ from one cohort to another. No differences were found in size of lymph nodes (14 mm vs 16 mm), solid (37 mm vs 34 mm) and cystic (30 mm vs 33 mm) pancreas lesions biopsied in each cohort (p > 0.05). PT in Cohort A underwent a lower number of EUS-FNA passes than PT in cohort B (1.5 vs 3.9; p < 0.05). Overall diagnostic accuracy in Cohort A and B was 90% and 83% (p = 0.05), respectively. (See table). Conclusions: 1.) On site EUS-FNA cytology assessment significantly improves diagnostic accuracy in solid pancreas masses, but not in lymphadenopathy or cystic lesions. 2.) On site cytology assessment reduces the number of passes required for diagnosis. Tabled 1 (∗p<0.05) Adenopathy (cohort A/B)(n = 50/50) SolidPancreas (cohort A/B)(n = 50/50) CysticPancreas (cohort A/B)(n = 50/50) Sensitivity 89%/84% 88%/77% 75%/74% Specificity 100%/100% 100%/100% 100%/100% Accuracy 94%/88% 93%/80%∗ 84%/83% Open table in a new tab