Accuracy of visual on-site evaluation (Vose) In predicting the adequacy of Eus-guided fine needle biopsy: a single center prospective study

Objective: Endoscopic ultrasound is the standard procedure for the diagnosis of pancreatic lesions and new needles have been developed to improve tissue acquisition (FNB). Rapid onset evaluation (ROSE) decreases the number of needle passes but is not always available. We introduced an easy and rapid method of direct classification of EUS-FNB sample namely Visual on-site evaluation (VOSE). Aims: To assess the accuracy of VOSE in predicting the histological adequacy of specimens. To evaluate the diagnostic power of FNB and the rate of core tissue obtained. Methods: Prospective single center study on patients with pancreatic lesions that underwent EUS-FNB. VOSE parameters were presence of blood, macroscopic visible core (MVC), number, color and length of specimen. The association between VOSE tool and histological adequacy was assessed. Fisher's exact test and Student's t-test used to compare categorical and continuous variables. Logistic regression analysis was used to assess association between variables. Results: 99 patients (58.6% male; mean age 68.4 ± 10) enrolled, including 102 lesions. Total number of passes was 358 with median number of 4 (range, 2-4). The 92.7% of samples were adequate and it was higher with the 22-G needle than with 25G (96.5% vs 89.2% p 0.01). VOSE "red-mixed specimen" was associated with a higher probability of histological adequacy (OR 2.39 95% CI 1.03-5.42 p = 0.04). Conclusions: The VOSE tool "red-mixed specimen" can be used to predict the histological adequacy and guide the number of needle passes. Overall, FNB provides a high rate of adequate and diagnostic specimen and high rate of core tissue especially with the 22G needle.