This is a prospective and comparative study including 76 consecutive patients performing EUS-FNB for pancreatic and extrapancreatic solid lesions, randomized by alternate allocation to macroscopic on-site evaluation (MOSE) (40 patients) or to a conventional technique (40 patients), with three passes each. MOSE samples were differentiated into score 0: no visible material, score 1: only necrotic or haematic material, score 2: white core tissue ≤ 2 mm, or score 3: white core tissue > 2 mm. The conventional technique consisted in pushing all the needle content into a test tube for evaluation by the pathologist. In both groups, a 22-25 Gauge Franseen-tip needle (Acquire, Boston Scientific Co., Natick, MA, USA) was used. The study evaluated the diagnostic accuracy and adequacy of MOSE compared to the conventional technique and whether MOSE could optimize the number of passes during EUS-FNB. Results: The analysis was performed on 76 patients (38 MOSE, 38 conventional). The overall diagnostic adequacy was 94.7% (72/76) and accuracy was 84.2% (64/76). The diagnostic accuracy was similar in the two groups: MOSE 86.8% (33/38 lesions), vs. conventional 81.6%, 31/38 lesions, p = 0.76). Regarding diagnostic adequacy, the MOSE technique was 97.4% (111/114 passes) compared to 92.1% (105/114 passes) with the conventional technique, p = 0.06. The accuracy increased according to the MOSE score evaluation: it was 43.5%, 65.5% and 78.3% in patients with score 1, score 2, and score 3, respectively. Moreover, if in the first two passes the MOSE score was 2 or 3, the accuracy was 82.6% (20/23), and upon adding a third pass, the accuracy increased to 87% (20/23), which was not significantly different from the general accuracy of the MOSE samples (86.8%) (p = 0.86). Conclusions: The MOSE score showed a comparable diagnostic accuracy to the conventional technique. However, MOSE allows endoscopists to perform an inspective evaluation of the material, tends to perform better than the conventional technique in terms of diagnostic adequacy, and may potentially reduce the number of passes.
Keywords: EUS; FNB; MOSE.