内镜超声引导下细针穿刺活检（EUS-FNB）在胰胆管肿块组织诊断方面已多次取代细针穿刺。然而，对于诊断恶性肿瘤所需的最佳穿刺次数尚不清楚。我们旨在比较两种类型的FNB针头在检测恶性肿瘤方面的单次穿刺表现。共114名因固体胰胆管肿块而被推介进行EUS评估的患者进行随机分组，其中一组采用弗兰森针刺活检，另一组采用具有不对称切割面的三钩尖针刺活检。每个肿块采取四次FNB。两位不知道穿刺针类型的病理学家对标本进行了分析。根据FNB病理、手术或FNB后至少6个月的随访，最终诊断是否为恶性肿瘤。比较两组FNB诊断恶性肿瘤的敏感性。计算每个臂的每次穿刺后EUS-FNB检测到恶性肿瘤的累计敏感性。还比较了样本的其他特征，包括细胞性和含血量。在主要分析中，将被归为可疑的病变不被视作诊断性恶性肿瘤。其中98名患者（86％）最终被诊断为恶性肿瘤，16名患者（14％）患有良性疾病。用弗兰森针采用四次EUS-FNB检测到47名患者中的44名患有恶性肿瘤（敏感性为93.6％；95％置信区间[CI]：82.5％-98.7％），而用三钩尖针在51名患者中检测到50名患的恶性肿瘤（敏感性为98％；95％ CI：89.6％-99.9％）（P：0.35）。用弗兰森针进行两次FNB检测到91.5％的患者有恶性肿瘤（95％置信区间[CI]：79.6％- 97.6％），而用三钩尖针检测到90.2％的患有恶性肿瘤（95％CI：78.6％-96.7％）。在第三次穿刺时的累计敏感性分别为93.6％（95％CI：82.5％-98.6％）和96.1％（95％CI：86.5％-99.5％）。用弗兰森针收集的样本细胞密度显著高于用三钩尖针收集的样本（P <0.01）。然而，在样本血性方面，两种针头没有差别。在疑有胰胆管癌的患者中，弗兰森针与三钩尖针的诊断表现没有显著差异。但是，弗兰森针产生的标本细胞密度更高。在使用任一类型的针头时，检测到病变至少需要两次FNB，敏感性可达90％以上。gov编号：NCT04975620。版权所有©2023年美国胃肠内窥镜学会。Elsevier Inc.发表。保留所有权利。

Endoscopic ultrasound-guided fine needle biopsy (EUS-FNB) has largely replaced fine needle aspiration for tissue diagnosis of pancreatobiliary mass lesions. However, the optimal number of passes required for the diagnosis of malignancy is not clear. We aimed to compare per-pass performance of two types of FNB needle for the detection of malignancy.Patients referred for EUS evaluation of solid pancreatobiliary mass lesions (n=114) underwent randomization between biopsy with a Franseen needle and a three-prong tip needle with asymmetric cutting surfaces. Four passes of FNB were taken from each mass lesion. Two pathologists blinded to the needle type analyzed the specimens. Final diagnosis of malignancy was made based on FNB pathology, surgery, or a follow up of at least 6 months after FNB. The sensitivity of FNB to diagnose malignancy was compared between the two groups. The cumulative sensitivity of detection of malignancy by EUS-FNB was calculated after each pass in each arm. Other characteristics of the specimens including cellularity, and blood contents were also compared between the two groups. In the primary analysis, lesions categorized as suspicious on FNB were considered non-diagnostic for malignancy.Ninety eight patients (86%) had a final diagnosis of malignancy, and 16 patients (14%) had benign disease. Four passes of EUS-FNB with the Franseen needle detected malignancy in 44 of 47 patients (sensitivity of 93.6%; 95% confidence interval [CI]: 82.5%-98.7%), and with the 3-prong asymmetric tip needle in 50 of 51 patients (sensitivity of 98%; 95% CI: 89.6%-99.9%) (P: 0.35). Two passes of FNB detected malignancy with 91.5% sensitivity (95% confidence interval [CI]: 79.6%-97.6%) with the Franseen needle, and 90.2% (95% CI: 78.6%-96.7%) with the 3-prong asymmetric tip needle. The cumulative sensitivities at pass 3 were 93.6% (95% CI: 82.5%-98.6%), and 96.1% (95% CI: 86.5%-99.5%), respectively. Samples collected with the Franseen needle had significantly higher cellularity than samples collected with the 3-prong asymmetric tip needle (P<0.01). However, there was no difference between the two types of needles in term of specimen bloodiness.There was no significant difference in the diagnostic performance of the Franseen needle versus the 3-prong asymmetric tip needle in patients with suspected pancreatobiliary cancer. However, the Franseen needle yielded higher cellularity of the specimen. Two passes of FNB are required to detect malignancy with at least 90% sensitivity with either type of needle.gov number: NCT04975620.Copyright © 2023 American Society for Gastrointestinal Endoscopy. Published by Elsevier Inc. All rights reserved.