Endoscopic Ultrasound-guided Fine-needle Biopsy for Tissue Sampling of Biliary Strictures: a Multicenter Prospective Study

After the disappointing results of first-generation reverse bevel fine-needle biopsy (FNB) needles, newer 'third generation' endoscopic ultrasound (EUS) needles were developed, known as the "end-cutting" needles. These needle types have shown excellent diagnostic accuracy (\>90%) in pancreatic masses and several other lesions, even without rapid on site evaluation (ROSE) and are not associated with increased adverse event incidence . The main design of end-cutting FNB needles are the Franseen type, which has three symmetrically distributed needle points and cutting edges, and the Fork-tip type which has 2 protruding asymmetrical sharp points and six distal cutting edges; however, other end-cutting needles are currently available such as the three-prong asymmetric tip needle and the Menghini tip needle. Given the significant increase in diagnostic accuracy and sensitivity observed in other abdominal lesions, these needles are expected to improve the diagnostic outcomes of EUS-FNB of "pure" biliary strictures although clinical data are still lacking. The aim of this study is to ascertain the impact of end-cutting FNB needles in the diagnostic algorithm of biliary strictures in absence of a clear pancreatic or abdominal mass. This is a multicenter international observational prospective study. Consecutive patients diagnosed with biliary strictures in absence of a pancreatic or abdominal mass will be assessed for eligibility. Patients with jaundice requiring decompression, will undergo endoscopic retrograde cholangiopancreatography (ERCP) with tissue sampling (brushing and/or biopsy) and biliary drainage, preferably in the same session immediately after EUS-FNB. Endoscopic ultrasound will be performed using a curvilinear array echoendoscope. FNB with 22G or 25G end-cutting needles will be performed. The size of the needles will be based on the personal choice of the endoscopist. Likewise, the sampling technique, whether using the "slow-pull", the "dry suction", or the "wet suction", will be based on the physician's choice. Three passes will be performed as per current guidelines. Specimens will be collected in 3 vials to allow for analysis according to needle pass; one for the first pass, one for the second pass, and one for the third and any eventual subsequent passes. The ROSE will not be available in this study. Primary outcome will be diagnostic sensitivity of EUS-FNB and 45 patients will be prospectically enrolled.

INTRODUCTION Tissue sampling to establish the nature of a stricturing lesion represents a key step in the diagnostic work-up of biliary strictures. As already known, a significant proportion (70%-80%) of biliary strictures are malignant, higher in patients with a definitive mass lesion than in those without a mass on cross-sectional imaging (PMID: 23340062). Early diagnosis is thus important for directing patients to the proper therapeutic strategy. While transpapillary brush cytology or forceps biopsy after sphincterotomy during endoscopic retrograde cholangiopancreatography (ERCP) have been the standard sampling modality, endoscopic ultrasound (EUS)-tissue acquisition (TA) is a useful later addition to the diagnostic armamentarium. The pooled sensitivity of brush cytology and intraductal biopsies has been found to be 45% and 48.1% respectively, while the combination of the two only modestly increased the sensitivity to 59.4% (PMID: 25440678). On the contrary, EUS-TA has been reported to have a pooled diagnostic sensitivity of 83% for distal biliary strictures (PMID: 26422979). Moreover, the choice of the modality adopted would depend not only on the availability of expertise, but also on the location of the lesion and presence of jaundice. Indeed, EUS-TA has shown better sensitivity for distal lesions compared to proximal lesions (PMID: 26422979). The performance of EUS-TA in peri-hilar strictures showed less favourable results. A 2020 prospective evaluation of EUS-fine-needle aspiration (FNA) in the diagnosis of biliary strictures (n=97, 46% hilar) showed an overall sensitivity of 75% (CI 64%-84%). Subgroup sensitivity and negative predictive value (NPV) reached 95% and 93% for distal lesions without stenting, whereas for the group of stented peri-hilar lesions this was much lower at 56% and 33% respectively (PMID: 33140008). However, data on EUS-TA of biliary strictures are based only to EUS-FNA studies whereas data on EUS-fine-needle biopsy (FNB) are lacking in this setting. In the last three decades, the evolution of needles has undergone a remarkable change and is constantly under development, by designing innovative tip designs to allow for larger histological specimens with preserved tissue architecture, along with use of tissue for genomic profiling and organoids development that are very important in the era of oncological personalized medicine. As a result, the current era has seen a decisive shift from FNA to FNB needles. EUS-FNA needles have limitations that are well known: their diagnostic performance is dependent on a cytopathologist to render a rapid on-site evaluation (ROSE), no provision of a histological core tissue is possible, and insufficient tissue for risk stratification to tailor anti-cancer therapy. After the disappointing results of first-generation reverse bevel FNB needles, newer 'third generation' EUS needles were developed, known as the "end-cutting" needles. These needle types have shown excellent diagnostic accuracy (\>90%) in pancreatic masses and several other lesions, even without ROSE and are not associated with increased adverse event incidence (PMID: 37591258; PMID: 36990125; PMID: 35124072). The main design of end-cutting FNB needles are the Franseen type, which has three symmetrically distributed needle points and cutting edges, and the Fork-tip type which has 2 protruding asymmetrical sharp points and six distal cutting edges; however, other end-cutting needles are currently available such as the three-prong asymmetric tip needle and the Menghini tip needle. Given the significant increase in diagnostic accuracy and sensitivity observed in other abdominal lesions, these needles are expected to improve the diagnostic outcomes of EUS-FNB of "pure" biliary strictures although clinical data are still lacking. The aim of this study is to ascertain the impact of end-cutting FNB needles in the diagnostic algorithm of biliary strictures in absence of a clear pancreatic or abdominal mass. STUDY DESIGN Multicenter international observational prospective study. This study will be conducted according to the principles and the recommendations of the 2013 Declaration of Helsinki. STUDY PATIENTS Consecutive patients diagnosed with biliary strictures in absence of a pancreatic or abdominal mass will be assessed for eligibility. Diagnosis of biliary strictures will be performed according to current guidelines (PMID: 37307900). Patients with jaundice requiring decompression, will undergo ERCP with tissue sampling (brushing and/or biopsy) and biliary drainage, preferably in the same session immediately after EUS-FNB. Inclusion criteria * Age ≥18 years * Suspicious biliary stricture in absence of an abdominal mass at cross-sectional imaging * Distal and peri-hilar biliary strictures * Informed consent provided by the patient or the closest relative Exclusion criteria * Presence of a pancreatic or abdominal mass structuring the common bile duct * EUS not feasible * Known bleeding disorder that cannot be sufficiently corrected with fresh frozen plasma (FFP) * Use of anticoagulants that cannot be discontinued * INR \>1.5 or platelet count \<50.000 * Pregnancy PROCEDURAL TECHNIQUE Endoscopic ultrasound will be performed using a curvilinear array echoendoscope. FNB with 22G or 25G end-cutting needles will be performed. The size of the needles will be based on the personal choice of the endoscopist. Likewise, the sampling technique, whether using the "slow-pull", the "dry suction", or the "wet suction", will be based on the physician's choice. Three passes will be performed as per current guidelines. Specimens will be collected in 3 vials to allow for analysis according to needle pass; one for the first pass, one for the second pass, and one for the third and any eventual subsequent passes. The rapid-on site cytological evaluation (ROSE) will not be available in this study. ENDPOINTS Primary Endpoint Primary outcome will be diagnostic sensitivity of EUS-FNB, defined as true positives/true positives + false negatives. Secondary Endpoints * Sample adequacy, defined as the ability to procure histological samples adequate for diagnostic interpretation. A histologic specimen will be defined as an architecturally intact piece of tissue from the target lesion (PMID: 29074447). * Diagnostic accuracy (defined as true positive + true negative/ total number of patients). The gold standard for the final diagnosis will be the pathological report in the subset of patients undergoing surgery and the clinical follow-up for 6 months in non-surgical patients. * Diagnostic specificity, defined as true negatives/true negatives+false positives * Histological quality of the sample quantified using a score for tissue integrity and blood contamination of specimens validated elsewhere (PMID: 32433914) * Yield of each needle pass * Number of needle passes * Adverse events, classified according to ASGE lexicon (PMID: 20189503). Postprocedural AEs will be investigated during follow-up with phone calls at 2 weeks after the procedure. ENROLLMENT AND DROP-OUT Enrollment will be competitive between centers and will extend for a maximum period of 1 year after local Ethic Committee/IRB approval. Patients will be considered dropped-out in case of consent withdrawn or if lost before the end of follow-up. SAMPLE SIZE CALCULATION AND STATISTICAL CONSIDERATIONS On the basis of a previous study (PMID: 33140008) reporting 75% sensitivity with EUS-FNA of biliary strictures, the study is designed to detect an increase in diagnostic yield with end-cutting FNB needles to 94.8%, as reported in a previous meta-analysis (PMID: 31579703). Therefore, 45 patients will be required to have a 90% power to detect the target difference at a 0.025 significance level (adjusted for multiplicity considering the planned subgroup analysis between peri-hilar and distal biliary strictures) and expecting a drop-out rate of 10%. A 2:1 stratification according to the location of the biliary strictures will be considered; therefore, we expect to enroll 30 patients with distal biliary strictures and 15 patients with peri-hilar biliary strictures. DATA COLLECTED General information * Age (years) * Sex (M/F) * Charlson comorbidity index * BMI Lesion information * Location * Imaging studies performed * Lesion size (mm) * Main pancreatic duct (MPD) dilation * CBD caliber (mm) Procedural information * Sampling route (stomach, duodenum) * Number of passes * Sampling technique (dry suction, wet suction, slow pull) Outcomes * Adequacy at each pass * Sensitivity * Specificity * Accuracy * Quality of sample * Surgery * AEs * Final diagnosis