Recent advancements in puncture needles for endoscopic ultrasound-guided biopsy have led to the development of FNB needles, which are superior to conventional FNA needles in terms of tissue sampling capacity. The commonly used FNB needles include EchoTip ProCore™ (Cook Medical, United States), Acquire™ (Boston Scientific, United States), and SharkCore™ (Covidien-Medtronic Inc, United States). The EchoTip ProCore™ features a reverse-bevel design, the Acquire™ needle is characterized by a Franseen-tip, and the SharkCore™ is designed with a Fork-tip. The optimal needle design for sampling is still under debate. However, a meta-analysis comparing the diagnostic performance of end-cutting fine-needle biopsy needles in EUS-guided sampling of solid pancreatic masses reported that both Franseen-tip and Fork-tip needles significantly outperformed reverse-bevel and FNA needles regarding diagnostic accuracy and sample adequacy [19].
In the context of EUS-LB, studies comparing FNA and FNB needles have demonstrated the superior tissue yield of FNB needles [14, 15], thereby highlighting the utility of FNB needles in liver disease diagnostics. Among the FNB needles, the Franseen-tip and Fork-tip needles are predominantly used in EUS-LB. Nieto et al. [20] and Aggarwal et al. [21] reported a significantly higher specimen length in the Franseen-tip group in their retrospective and prospective studies, respectively. Conversely, Hashimoto et al. [22] found no significant differences in post-fix aggregate specimen length and post-fix longest specimen length between the two FNB needles in a repeated-measure crossover study with a prospectively maintained cohort of patients.
Our study is a single-center retrospective comparison of the Franseen-tip and Fork-tip needles, focusing on their efficacy and safety. A notable distinction from existing reports is the method of liver tissue yield assessment, which we verified by tissue surface area rather than total specimen length (TSL). The rationale behind this is the thinner nature of specimens obtained by EUS-LB compared to percutaneous liver biopsy (PLB), making length-based evaluation less reliable, especially considering the tendency of EUS-LB specimens to fragment. We previously reported that liver tissues obtained by both PLB and EUS-LB techniques with different gauge numbers, when evaluated in terms of surface area, yielded comparable tissue volume [10]. Therefore, we propose that surface area evaluation, facilitated by whole-slide imaging, represents a more accurate method for assessing EUS-LB tissue yield than TSL.
In this study, we performed univariate and multivariate analyses to explore the factors influencing liver tissue yield. Both analyses identified staging of F2 or greater and the presence of NAFLD as significant factors impacting the total tissue surface area. Additionally, platelet counts emerged as a significant factor in determining the surface area of the most tissues sampled per pass in multivariate analysis. These findings suggest that liver tissue volume tends to be lower in patients with severe fibrosis and higher in patients with NAFLD. The reduced tissue yield in livers with advanced fibrosis could be attributed to increased puncture resistance due to the needle's need to cut through dense collagen fibers, which hampers the needle's ability to encapsulate sufficient liver tissue.
Interestingly, our results indicate that EUS-FNB may procure more tissue in NAFLD patients compared to patients with other liver diseases. However, a bivariate analysis examining the relationship between fibrosis and NAFLD, specifically assessing fibrosis staging and platelet count, did not reveal a statistically significant difference, indicating that NAFLD cases did not exhibit significantly less fibrosis. This finding contrasts with Nieto et al. [20], who showed longer TSL associated with advanced fibrosis (F3-4) in both FNB needles. These disparate findings underscore the need for a standardized method of assessment, with our study suggesting that surface area evaluation through whole-slide imaging may provide a more accurate measure than TSL.
No studies to date have assessed liver tissue yield across different background liver diseases. The higher tissue yield observed in NAFLD patients could be due to (1) the inherently easier collection of specimens from fatty livers, or (2) a relative increase in tissue yield in NAFLD cases due to a higher prevalence of advanced fibrosis in other liver diseases. Moreover, this study noted a higher frequency of autoimmune liver diseases and suspected cases in pathological diagnoses. In autoimmune liver diseases, fibrosis tends to manifest after prolonged chronic active hepatitis, potentially leading to advanced fibrosis by the time of diagnosis. Although not assessed in this study, disease duration may influence liver stiffness and consequently affect biopsy results, suggesting it as a factor worth considering in future studies.
Our study is not without limitations. Being a single-center study with a relatively small sample size and retrospective in nature, it does not specify the disease of interest. Furthermore, the study lacks a comparison with normal liver tissue, providing limited evidence to support the hypothesis that livers with less fibrosis yield more tissue. Additional studies focusing on specific diseases are warranted. The amount of tissue sampled in EUS-LB may vary with the operator's skill level; in our study, both a trainer and a trainee performed each procedure. Despite the retrospective nature of our study possibly limiting the capture of all adverse events, we believe that no serious adverse events were missed, as all patients were followed up 1–2 weeks after discharge for test result collection.
In conclusion, the findings of this study indicate no significant differences in the surface area of liver tissue and the number of portal vein areas obtained using two distinct puncture needles. Both needle types successfully provided sufficient liver tissue for histopathological analysis. While previous studies [20, 21] have suggested superior tissue yield with the Franseen-tip needle, the current data do not conclusively establish the superiority of either needle. Consequently, a multicenter randomized controlled trial is warranted to definitively ascertain the superior needle type for tissue yield. Additionally, technical advancements and standardization of the EUS-LB procedure are highly desirable. It is hoped that the outcomes of this study will contribute to the refinement and standardization of the EUS-LB technique, enhancing its efficacy and reliability in clinical practice.