Clinical utility of macroscopic on-site evaluation in EUS-guided tissue acquisition for autoimmune pancreatitis: A retrospective study.

BACKGROUND AND OBJECTIVES: Pathological diagnosis of autoimmune pancreatitis (AIP) is essential for selecting the appropriate treatment strategy. Although macroscopic on-site evaluation (MOSE) can reduce the number of needle passes during EUS-guided tissue acquisition (EUS-TA) in pancreatic neoplasms, its utility in AIP remains unclarified. We investigated the utility of MOSE in diagnosing AIP using EUS-TA-derived specimens. METHODS: We conducted a retrospective study of patients with type 1 AIP who underwent EUS-TA using fine-needle biopsy needles. Macroscopic visible core (MVC) length was measured using MOSE, and histological diagnosis was evaluated following the International Consensus Diagnostic Criteria. The correlation between MVC length and diagnostic yield, histological features, and tissue volume was assessed. RESULTS: Thirty-eight lesions of 36 patients were included. Among these, 32 had definite AIP and four had probable AIP. Level 1 and 1 or 2 histologies were identified in 47% and 90% of lesions, respectively. Longer MVC length was associated with a higher level 1 histology detection rate (P < 0.01), with 22 mm as the optimal cutoff (area under the curve = 0.80; 95% confidence interval = 0.65-0.95). Specimens with MVC ≥22 mm (75%) had a significantly higher level 1 histology detection rate than those with MVC <22 mm (17%) (P < 0.001). MVC length positively correlated with tissue area (r = 0.789, P < 0.001) and independently affected level 1 histology detection (odds ratio = 11.2, P < 0.01). CONCLUSION: MVC length ≥22 mm is a practical indicator of adequate tissue quality for the histological diagnosis of AIP. Therefore, MOSE may help determine the number of needle passes during EUS-TA in patients with AIP.