Fine-needle aspiration (FNA), the erstwhile primary workhorse for endoscopic ultrasound (EUS) guided tissue acquisition of solid mass lesions, has now largely been replaced by fine-needle biopsy (FNB). Using FNB needles, a diagnostically adequate sample can be procured reliably in more than 90% of patients with the vast majority being histology-grade tissue. This has overcome several of the pitfalls encountered at EUS-FNA: clinical outcomes are not reliant on availability of rapid-onsite assessment, ancillary studies and molecular profiling can be more reliably performed on biopsy samples and an interpretation can be rendered by most gastrointestinal pathologists, not just cytopathologists. To further optimize procedural outcomes at EUS-FNB, recent studies have compared different FNB needle designs and evaluated several sampling methods. In a randomized trial that compared performances of the four most frequently used 22-gauge (G) FNB needle designs – Reverse-bevel, Menghini-tip, Fork-tip and Franseen-tip – it was observed that the Franseen-tip and Fork-tip needles yielded diagnostically adequate, histology-grade, samples in >90% of patients in a single pass.1 More importantly, the outcomes were excellent when suction was not applied or when stylet retraction was adopted during the sampling maneuver – these methods yielded a diagnostic adequacy of 94–100%. Additionally, a high cellularity sample, defined as ratio of core tissue to total specimen >50%, was achieved in 55–75% of patients when either of the two methods were adopted. On the contrary, while the application of suction yielded a diagnostic adequacy of 100%, high cellularity was achieved in only 35%. In another randomized trial that compared performances of the Franseen and Fork-tip needles in patients with solid pancreatic mass lesions, both needles were found to perform equally well, with diagnostic cell block exceeding 90% for both needle designs.2

A natural follow-up question to address in EUS-guided tissue acquisition is, do the 22G and 25G FNB needle designs perform equally well? At least for FNA, randomized trials have shown that the performance of the 22G and 25G needles are equivalent with no difference in diagnostic outcomes.3 However, given the thin caliber, 25G FNA needles appear to be more conducive for trans-duodenal tissue sampling with fewer rates of needle dysfunction and hence lowers procedural costs.4 To determine whether the gauge of an FNB needle has any impact on procedural outcomes, Tomoda et al. conducted a non-inferiority, cross-over, randomized study that compared the performance of 22G and 25G Franseen needles in 88 patients with solid pancreatic mass lesions.5 Diagnostically adequate, histology-grade, specimens were procured in 70% and 78% of the 25G and 22G FNB cohorts, respectively, indicating non-inferiority for performance of the 25G needle. Technically, the ease of puncture was superior for the 25G needle cohort, 69% versus 52%, with adjusted difference of 12.2%, P < 0.001. However, high-quality histology samples were obtained in 45.5% of the 22G as compared to 25% in the 25G cohort. Although this result did not show non-inferiority for 25G as compared to the 22G cohort (adjusted difference −22.7%; 95% CI −24.1% to −21.2%; P = 0.999), the superiority of 22G to 25G needles in procuring high-quality tissue was evident. All procedures in both cohorts were performed by applying suction during the sampling maneuver. Also, the non-inferiority margin was determined to be 15%, a number too high for clinical practicality.

The findings of the present study mirror the findings of another clinical trial from Indiana University that compared performances of the 22G and 25G Franseen needles in patients with solid pancreatic mass lesions.6 In that study, while the diagnostic yield between cohorts were similar (25G 98% vs. 22G 88%, P = 0.105), the 25G cohort required additional passes to procure adequate tissue for the cell block. Also, the authors implied that the 25G needle appeared more conducive for performing trans-duodenal sampling. In the present study by Tomoda et al., a subgroup analysis was not performed to identify the cohort of patients in whom the technical performance of the 25G needle was superior. A small sample size likely precluded this determination. While significantly better histology-grade tissue could be procured using 22G FNB needles in the study by Tomoda et al.,5 the Indiana study also suggests that the quantity of tissue procured by 25G needles may be less thereby necessitating additional passes.6 As alluded to earlier, in a recent randomized trial, our group has shown that high-quality histology-grade material can be procured reliably using the 22G Franseen needle in a single pass, provided suction is not applied during the sampling maneuver.1

How are we to interpret the findings of the present study and apply it to our clinical practice? In an ideal situation, patients should be subjected to a single EUS procedure that will (i) yield a definitive diagnosis, (ii) procure adequate, high-quality, histology-grade tissue for ancillary testing, and (iii) the above outcomes be achieved with least number of passes possible. Present evidence suggests that while both 22G and 25G needles perform equally well from a diagnosis perspective, high-quality, histology-grade, tissue is best procured with the larger 22G needle, and with fewer passes. In select patients, particularly those requiring trans-duodenal sampling in whom the mass may be small (<1–2 cm) or difficult (distal bile duct or uncinate lesion) to access, a 25G FNB needle may be optimal. In those challenging cases, it is imperative to perform additional passes to ensure sufficient tissue procurement for ancillary studies. However, more studies are needed to prove this assumption. It is also our preference that when using the 22G needle, the stylet retraction or no suction method be applied during the sampling maneuver to increase the likelihood of procuring high-quality histology samples.

J.Y.B. is a consultant for Boston Scientific, Olympus, Fujifilm Medical Systems. S.V. is a consultant for Boston Scientific, Olympus, Medtronic.