Pearls and Pitfalls of Real-Life Molecular Testing on FNA and Core Biopsy in Pancreatic Adenocarcinoma Practice.

Objectives The diagnostic efficacy of endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) of pancreatic solid masses varies widely depending on the presence of rapid onsite evaluation (ROSE) and the type of cytology preparations. The present study aims to compare the diagnostic efficacy of cytology smears (CS) and cell block (CB) with ROSE versus Thinprep slides (TpS) with CB without ROSE in diagnosing pancreatic solid masses. Methods Performed retroprospective review of 56 patients (2017-2018) with surgically confirmed pancreatic neoplasms including cytology preparation type and comparison of diagnostic efficacies. Results Out of 56 patients who underwent surgical resection, 30 patients were diagnosed with pancreatic ductal adenocarcinoma (PDAC) and 26 with pancreatic neuroendocrine tumor (PNET) or solid pseudopalliary neoplasm (SPN). Out of 30 surgically removed PDACs, 20 patients had prior cytology specimens while 10 had surgical core biopsies. Of the 20 cytology specimens, 12 patients had ROSE with concurrent CS and CB, and all cases were diagnosed as positive. Out of the 8 patients without ROSE, 3 patients (1 with bile duct brush, 2 with TpS and CB) were diagnosed as positive. The diagnostic efficacy was 100% (12/12) in patients with ROSE and CS compared to 37.5% (3/8) in patients without ROSE. In 26 non-PDAC cases, 21 patients had cytology specimens and 5 had surgical biopsies. Seventeen cases with ROSE and immunohistochemical stains (IHC) on CB were diagnosed as positive, while 4 cases without ROSE and IHC received atypical diagnoses, resulting in a decreased diagnostic efficacy from 100% (17/17) to 0% (0/4). Conclusion ROSE with CS and CB are superior to TpS and CB in diagnosing solid pancreatic masses. ROSE with IHC on CB is crucial for diagnosing PNET and SPN, while CS and CB are important for PDAC diagnosis.

Rapid on-site evaluation (ROSE) improves the diagnostic yield of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA). The cytopathologist performs direct visualization of the cytological smear by microscopy to confirm specimen adequacy and provides a preliminary diagnosis. Further sampling is performed immediately if the specimen is inadequate, thus avoiding the need for repeat procedures.1 ROSE can also reduce the number of unnecessary needle punctures during EUS-FNA, as the procedure is stopped once adequate tissue sample is obtained, potentially reducing the risk of local complications. ROSE is not widely available in many centres in Asia, due to logistical challenges of having a cytopathologist in attendance during all EUS-FNA procedures, and it also incurs additional manpower cost. In recent years, there have been significant advances in artificial intelligence (AI)-based systems, which use algorithms to perform tasks that would usually require human intelligence and input, and AI has now been incorporated into medical practice.2 AI-assisted colonoscopy is now part of regular clinical practice, and it has been shown to significantly improve adenoma detection rate.3 There is ongoing research into the use of AI in upper endoscopy, EUS and capsule endoscopy for lesion detection and characterization, as well as quality tracking.4 The introduction of digitalized whole slide image technology has enabled the application of digital pathology in clinical practice and education. This digital transformation has in turn facilitated the implementation of AI-based algorithms in the field of histopathology in the context of AI-driven cancer diagnostic, prognostic, and predictive applications.5 Published literature on the application of AI-models in the field of cytopathology are fewer in comparison, with a focus on gynecological samples initially. In a systematic review on recent application of AI in non-gynecological cancer cytopathology, only 28 published original studies with full texts from 2010 to 2021 were identified. These studies involved 8 target organs and only one study involved the pancreas.6 It is a natural progression to explore application of AI support systems during the EUS-FNA procedure. This represents the interface between endoscopy and cytopathology, in the context of real-time assessment of adequacy of tissue sample and providing a provisional diagnosis before formal histopathological assessment. In this issue of Journal of Gastroenterology and Hepatology, Lin et al. reported the validation of a novel AI-based model (ROSE-AI model) to substitute for ROSE during EUS-FNA.7 Digitized images from Diff-Quik-stained EUS-FNA slides from 51 patients, predominantly for pancreatic lesions (94.1%), were used to train and validate the AI-algorithm, with an emphasis on specificity for malignancy. The ROSE-AI model achieved an accuracy of 83.4% in the internal validation dataset and 88.7% in the external test dataset. The sensitivity and positive predictive value were 79.1% and 71.7% in internal validation dataset and 78.0% and 60.7% in external test dataset, respectively. The specificity and negative predictive value were 85.4% and 89.7% in the internal validation dataset and 90.6% and 95.7% in the external test dataset, respectively. The accuracy for the training dataset of this AI-ROSE model is 97.7%, which is comparable to the model described in an earlier non-validated study involving 75 FNA pancreas samples that highlights its potential in improving the diagnosis for cases that fall within the atypical diagnostic category.8 The study by Lin et al. is one of the few to evaluate the application of AI-model in diagnostic cytopathology in the context of samples obtained from EUS-FNA of pancreatic lesions, and the first to demonstrate the potential feasibility of using an AI-system to replace manual ROSE during EUS-FNA. The study design was appropriate with the appropriate method of algorithm development in terms of training, internal and then external validation. However, with a sensitivity performance of under 80% in both the internal and external validation datasets, the results are still suboptimal, and this study remains a proof of concept. Additional issues also need to be addressed before it can be translated into clinical practice. When EUS-FNA is performed, the key issues are firstly whether adequate lesional tissue sample has been obtained, and secondly to establish if the acquired sample is truly benign or malignant. The sole focus of the current algorithm is on detection of malignant cytology. To replicate ROSE more closely, the ideal AI-model should be able to assess adequacy of sample acquired especially in the setting of blood contamination and provide a ‘benign’ or ‘malignant’ diagnosis with high confidence. In the current model, there is the need for high-quality staining, slide scanning and digitalization before running the machine-learning algorithm. These requirements present additional practical and financial obstacles for infrastructure development within the endoscopy unit and/or pathology laboratory. Digital pathology also has inherent disadvantages of additional workflows, need for additional personnel, and need for data storage space, which further increase the operational cost and procedure time. During ROSE, the experienced cytopathologist performs microscopic examination on freshly prepared slides without coverslips when they are often still wet to achieve adequacy check and provide a preliminary diagnosis usually under 5 minutes. In their discussion, the authors made a general statement about the AI-ROSE model improving EUS-FNA efficiency without taking into consideration the additional time required for drying (slides must be fully dried before loading into the scanner to avoid residue getting onto the scanner mechanism) and cover-slipping the Diff-Quik-stained slides, slide scanning and digitalization prior to the process of running the machine-learning algorithm. This extra operational time may not be practical for the endoscopist and may potentially affect the safety of the patient who is under sedation. The significant cost in setting up a digital pathology system has limited its widespread adoption beyond well-funded academic medical centres where access to ROSE is likely to be available already. The ideal AI-ROSE support system should require less set-up costs and infrastructure support and should not significantly increase the procedure time. One possibility may well be building such AI-algorithms into portable devices, such as smartphones, that can simply make prediction on sample adequacy and presence of malignant cells by capturing and examining images from Diff-Quik-stained EUS-FNA slides. Regarding challenges about medicolegal responsibility and regulatory issues of an interim report by an AI-ROSE model, another more practical solution is engaging an off-site cytopathologist by way of telepathology with the use of digital pathology. Current alternative strategies to ensure specimen adequacy include macroscopic on-site quality evaluation (MOSE) by the endoscopist to ensure at least 4 mm of macroscopic visible core tissue,9 and indirect methods without tissue visualization, such as performing a pre-determined minimum number of needle passes, which has been correlated with obtaining adequate tissue sample,10 and the use of biopsy needles, which can increase the amount of core tissue.11 Although MOSE is useful, it does not directly assess whether the lesional tissue was sampled, and not adjacent normal tissue, and it just provides an indication of adequacy of tissue sample for further analysis. This holds true for the two indirect approaches in that they are surrogates to ensure adequate tissue has been obtained and do not differentiate between lesional or non-lesional tissue and cannot provide a provisional diagnosis real-time. However, all these approaches have served us well in actual clinical practice and even as we seek to explore new solutions, we can still rely on them in our practice. Cytology smears alone may be insufficient for complete diagnosis. Additional material may be required for ancillary immunohistochemistry and molecular studies for diagnosis and personalized therapeutics. Nonetheless, development of the AI-ROSE model is a positive step forward and when refined and available for clinical practice, it will further expand our armamentarium to improve the diagnostic yield of EUS-guided tissue acquisition.

The study aimed to demonstrate rapid and effective molecular testing on liquid-based cytology (LBC) samples for EGFR, KRAS and BRAF mutations using the Biocartis Idylla™. Rapid on-site evaluation (ROSE) LBC samples for patients with non-small cell lung carcinoma (NSCLC) or pancreatic ductal adenocarcinoma (PDAC) were tested for EGFR, KRAS and BRAF mutations based on the relevance to tumour subtype. The quantification values (Cq values) and mutation detection status were compared between LBC samples and routine formalin-fixed paraffin-embedded (FFPE) clot samples. ROSE LBC samples (n = 54) showed a higher yield of well-preserved tumour and wild type (WT) DNA, demonstrated by lower quantification cycles, no false positives or false negatives, and a higher sensitivity for low allele frequency mutations when compared with FFPE clot samples. The Biocartis Idylla™ provides highly sensitive, reliable and rapid testing for LBC samples for the detection of EFGR and KRAS mutations. BRAF mutations were not detected in the participant cohort; however, all LBC WT BRAF results correlated with the results from the FFPE clot samples. Access to rapid molecular testing using LBC samples can detect the most frequent driver mutations closer to the time of diagnosis, enabling the selection of the most effective first-line targeted therapy sooner, reducing delays or side effects from suboptimal treatments, patient anxiety and costs to healthcare systems, whilst improving patient outcomes.

Transesophageal Endoscopic Ultrasound-Guided Mediastinal Lymph Node Aspiration: Does the End Justify the Means?

Background: Endoscopic ultrasound-guided tissue acquisition (EUS-TA) is a well-established diagnostic tool to obtain tissue from solid pancreatic masses, gastrointestinal subepithelial lesions, lymph nodes, liver as well as other neighboring organs. EUS-TA has proven to have a major clinical impact in diagnosis and staging, decreasing the need for invasive diagnostic procedures, such as thoracoscopy, mediastinoscopy, laparoscopy and open surgery. EUS-TA includes endoscopic ultrasound guided fine needle aspiration (EUS-FNA) and endoscopic ultrasound guided fine-needle biopsy (EUS-FNB). Both procedures are safe and yield high diagnostic value. Despite its high diagnostic yield, EUS-FNA limitations include the inability to determine histologic architecture, and a small quantitative sample which may be inadequate for further immunohistochemical staining or molecular testing. EUS-FNB, with its larger core biopsy needle, was designed to overcome these potential limitations. It remains unclear if EUS-FNB has truly overcome these obstacles, and the optimal selection of FNA or FNB in different clinical contexts for different lesions is yet to be determined. We present a review on studies examining EUS-FNA vs EUS-FNB. Conclusion: For solid pancreatic masses, there is no difference in diagnostic accuracy or tissue cores rates when accompanied by rapid on-site evaluation (ROSE). The diagnostic yield is higher in FNB compared to FNA in cases where ROSE is not accessible. In addition, one study indicated that the combined approach of EUS-FNA + FNB resulted in a higher diagnostic yield than EUS-FNA alone, accompanied by a reduced number of needle passes required. In solid gastrointestinal lesions, FNB is associated with a relatively better diagnostic adequacy and tissue cores rates, with less number of needle passes. Regarding lymph node biopsy, the availability of prospective trials is limited. Based on the current literature, we would like to propose EUS-FNB as the recommended approach for lymph nodes.

BackgroundEndoscopic ultrasound-guided fine needle aspiration (EUS-FNA) technology is widely used for the diagnosis of pancreatic masses. However, in some cases, inadequate tissue volume or difficulty of morphological diagnosis are constraining factors for adequate cytopathological evaluation. K-ras mutation is the most frequently acquired genetic abnormality, occurring in approximately 90% of all patients with pancreatic ductal adenocarcinoma (PDAC). In the present study, the clinical utility of residual liquid-based cytology (LBC) specimens obtained using EUS-FNA for K-ras mutation analysis was evaluated.MethodsIn this study, 81 patients with pancreatic lesions were examined. The cell block (CB) specimens separated from EUS-FNA samples were morphologically evaluated by hematoxylin–eosin (HE) staining. Final diagnoses were confirmed by CB specimens, surgical resection specimens, diagnostic imaging, and clinical follow-up. Genomic DNA of residual LBC specimens stored at 4°C for several months were extracted and assessed for K-ras mutations using a fluorescence resonance energy transfer-based preferential homoduplex formation assay.ResultsK-ras mutation analysis using residual LBC samples was successful in all cases. The sensitivity, specificity, and accuracy of CB examination alone were 77.4%, 100%, and 81.3%, respectively, and those of the combination of CB examination and K-ras mutation analysis were 90.3%, 92.3%, and 90.7%, respectively. Furthermore, K-ras mutations were detected in 8 (57.1%) of 14 PDAC samples for which the CB results were inconclusive.ConclusionThese findings suggest that K-ras mutation analysis using residual LBC specimens improves the diagnostic accuracy of EUS-FNA.

Now that tissue cores can be obtained using fine-needle biopsy (FNB) needles, the ways tissues are handled for endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) are changing. Direct smear, touch smear of core tissues, and centrifugation have been used for cytological examinations, and liquid-based cytology (LBC), which allows immunostaining and genetic tests that use residual samples, is emerging as an alternative. We emphasize that liquid cytology (Cytospin™ cytology and LBC) is still important, because it enables the diagnosis of pancreatic ductal adenocarcinoma (PDAC) when cancerous cells are scarce in specimens. Cell blocks are being replaced by core tissues obtained via FNB needles. Recent reports indicate that rapid on-site evaluation (ROSE) is not necessary when FNB needles are used, and macroscopic on-site evaluation is used to evaluate specimen adequacy. Macroscopic findings of specimens are helpful in the diagnostic workup and for clarifying specimen-handling methods. In addition to the red strings and white cores observed in PDAC, mixed red and white strings, gray tissues, and gelatinous tissues are observed. Gray (necrotic) tissues and gelatinous (mucus) tissues are more suitable than histology for cell block or cytological processing. Tumor cells in neuroendocrine tumors (NETs) are numerous in red strings but cannot be observed macroscopically. ROSE might thus be necessary for lesions that may be NETs. Core tissues can be used for genetic tests, such as those used for KRAS mutations and comprehensive genomic profiling. Cytological materials, including slides and LBC specimens, can also be genetic test materials.

Introduction: Currently, there is a lack of reliable markers useable in diagnosis and/or management of the pancreatic ductal adenocarcinoma (PDAC). Although detection of KRAS point mutations for the differential diagnosis and determination of miR-21 expression for diagnostic and prognostic purposes have been widely investigated with promising results, neither KRAS nor miR-21 testing have become routine in clinical practice so far. One of the reasons is that analysis of KRAS mutations and miRNA expression is mainly performed on resected pancreatic tissue the use of which is relevant only for a minority of PDAC patients undergoing of surgical treatment. The present study describes development of a reliable methodology for analysis of KRAS mutations and miR-21 expression from samples acquired by endoscopic ultrasound-guided fine-needle biopsy (EUS-FNB), applicable to all PDAC patients. Consequently, the success rates and clinical validity of KRAS mutation and miR-21 expression analysis in two common types of EUS-FNB samples, i.e. native aspirates and cytology specimens are compared. Experimental: The study included 118 patients with a confirmed diagnosis of pancreatic ductal adenocarcinoma (PDAC). Each patient underwent EUS-FNB and the sample was divided into two parts, one part was stored in a stabilizing solution as native aspirate and the rest was processed into the cytology specimen. DNA/RNA was extracted and then analyzed for KRAS mutations and miR-21 expression. For both sample types, the yields of DNA/RNA and success rates for KRAS and miR-21 testing were compared along with evaluation of mutant cell fractions. Finally, confirmation of miR-21 prognostic role was tested by Kaplan-Meier analysis. Data Summary: The overall amount of isolated DNA/RNA from native aspirates was significantly lower compared to the cytology specimens (147 ng vs. 642 ng for DNA, 10 ng vs 164 ng for RNA). The success rates for subsequent KRAS and miR-21 analysis was 100% for both sample types. The KRAS-mutant detection rates in native aspirates were 12% lower than in cytology specimens (78% vs. 90%). The average fraction of KRAS-mutant cells was lower in native aspirates (31%) compared to the cytology specimens (56%). The prognostic role of miR-21 in native aspirates did not reach statistical significance (p = 0.06), but was confirmed in cytology specimens (p = 0.02). Conclusions: Although both types of EUS-FNB samples are suitable for DNA/RNA extraction and subsequent DNA mutation and miRNA expression analysis, reliable results with clinical validity were only obtained for cytological specimens. Native aspirates exhibit a low and undefined fraction of tumor cells leading to false interpretation of molecular testing results. EUS-FNB samples in the form of cytology specimens are a perspective source for the study of molecular markers in virtually all PDAC patients, including patients in inoperable stages. Supported by Internal Grant Agency of the Czech Ministry of Health (IGA) grant No. NT 13638 Citation Format: Lucie Benesova, Tereza Halkova, Bohus Bunganic, Barbora Belsanova, Eva Traboulsi, Miroslav Zavoral, Marek Minarik.{Authors}. Molecular cancer testing of KRAS and miR-21 from EUS-guided biopsies of pancreatic tissue: Utility of aspirates vs. cytology. [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2016 May 12-15; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2016;76(24 Suppl):Abstract nr B18.

Endoscopic ultrasonography

Non-Diagnostic EUS-FNA in Patients With Suspected Pancreatic Adenocarcinoma: Risk Factors, Repeat Sampling, and Impact on Diagnosis and Treatment

EUS-guided fine-needle biopsy sampling versus FNA in the diagnosis of subepithelial lesions: a large multicenter study

Analysis of circulating cell-free DNA after endoscopic ultrasound-guided fine needle aspiration in pancreatic ductal adenocarcinoma

Endoscopic ultrasound guided fine needle aspiration (EUS FNA) has been in existence for more than 20 years with the first EUS FNA performed in 1992 1 followed by first EUS FNA diagnosis of pancreatic cancer in 1994 2. Since then, it has proven to be a very safe and reliable procedure. Multiple studies have shown an impeccable safety record. In a recent study of 3090 consecutive EUS FNA patients, the bleeding risk and perforation risk was only 0.1% and 0.01%, respectively 3. A systematic review reported an overall morbidity of 0.98%, pancreatitis of 0.44%, post-procedural pain of 0.34% and mortality of 0.02% in 10 941 patients undergoing EUS FNA 4. It is an indispensable modality in obtaining tissue diagnosis from various solid and hollow viscera as well as lymph nodes. In particular, pancreatic mass lesions are diagnosed with high sensitivity and specificity. Systematic review and meta-analysis have reported a sensitivity of 85.0–86.8% and specificity of 95.8–98.0% 5, 6. The sensitivity and specificity for differentiating malignant from benign lesions was 91.5% and 97.7% 7, respectively. Even for small lesions measuring less than 1 cm, the diagnostic accuracy is 96% 8. Diagnosis of intramural lesions within a hollow viscous was initially more challenging. Williams et al. published an accuracy of 50%, sensitivity of 25% and specificity of 38% in 1999 9. However, subsequent studies have demonstrated higher diagnostic accuracy between 80.0% to 95.6% 10-13. EUS FNA of lymph nodes is frequently successful with adequate specimens obtained in 87% of cases with sensitivity, specificity and diagnostic accuracy between 71–75%, 95–99% and 85–90%, respectively 14, 15. When considering mediastinal lymph node sampling, a meta-analysis of 76 studies showed a sensitivity of 88% and specificity of 96% 16. Despite the high success rate of EUS FNA in securing a diagnosis with a minimal complication rate, there is still potential for refinement influenced by choice of needle, FNA technique augmentation as well as specimen handling to further improve clinical outcomes for patients. There is currently a wide range of needle sizes and types at one's disposal. Needle size options include 19, 22 and 25 gauges, each possessing unique advantages and disadvantages. Needle types include: straight needle for FNA, spring-loaded Trucut and reverse side bevel for FNA/B, and other emerging novel needles (e.g. built in side port, cryotechnology). Larger needles theoretically acquire a larger volume of tissue; however, it does not always equal a higher yield 17 as technical factors may come into play and conversely a smaller needle may be advantageous. In general, the size of needle should be tailored towards what the endosonographer is aiming to achieve (cytology vs histology) and characteristics of the target lesion. Thus, the needle size and type is tailored to the specific lesion targeted and clinical context. The 19 g needle (Fig. 1) is useful in situations where a larger volume of cells or core histology 12, 18-21 is required. For example, the large needle may be used to make a diagnosis of autoimmune pancreatitis 22 or for the diagnosis of lymphoma and certain neuroendocrine tumors, where immunohistochemical staining is needed. Although more tissue can be potentially obtained with the 19 g, this can be hampered by technical failures due to stiffness of the needle and an angulated or torqued scope position resulting in reduced maneuverability compared to the 22 g 23, 24. A modified technique has been described involving removal of the stylet prior to extension of the 19 g needle in the duodenal lumen to improve technical success in the pancreatic head and uncinate areas, which had an overall feasibility of 99% 12. However, the proportion of lesions requiring transduodenal puncture was not entirely clear and 62.5% of cases used a forward viewing echoendoscope, which has a larger working channel and no elevator, and therefore, may not be comparable to the standard linear echoendoscope. Nevertheless, a flexible 19 g nitinol needle (Expect 19 Flex; Boston Scientific, Natick, MA, USA) had been developed and 100% technical success for FNA of lesions around the head and uncinate have been reported 20, 25. When compared to the smallest needle (25 g), a recent multicenter randomized control trial (RCT) published in abstract form reported no significant difference in diagnostic accuracy when targeting large pancreatic lesions. The only significant difference between the 19 g and 25 g was greater bloody contaminant in the 19 g group 26. Currently most operators have the 22 g or 25 g (Fig. 1) as their needle of choice when performing FNA in solid lesions. The advantage of a smaller caliber 25 g needle in general include maneuverability during transduodenal FNA 17 whilst obtaining high quality cellularity with less bloody contamination by utilizing its innate capillary action, which negates the need for syringe suction and may be theoretically safer. There have been multiple studies comparing the 22 g vs 25 g straight needles showing similar overall diagnostic yield 27-30. One study showed a non-statistically significant difference in diagnostic yield of 87.5% (22 g) vs 95.5% (25 g) (P = 0.18) both with similar number of passes and no complications 28. Another study with 842 patients showed the sensitivity and specificity of the 22 g and 25 g was 84% and 100% vs 92% and 97% respectively; however, the P-value was not disclosed. There was a 2% rate of pancreatitis in the 22 g group and none in the 25 g group 29. A recent systematic review and meta-analysis comparing 22 and 25 g, which showed better specimen adequacy in the 25 g with similar diagnostic accuracy and complication rates 30. Lastly a meta-analysis involving 1292 patients showed the 25 g was more sensitive compared to the 22 g (93% vs 85%, P = 0.0003) but comparable specificity (100% vs 97%) 31. On balance, the 25 g FNA needle appears to be the needle of choice for a cytologic diagnosis, particularly for pancreatic solid lesions, unless core histology is required or ancillary testing such as molecular markers or special staining are needed. Initially only the straight tipped needle was available. This was capable of high cytological yield; however, in situations where core histology is required, outcomes were suboptimal 23. This led to alternative needle designs in pursuit of a higher histological yield, especially, as core specimen have become increasingly important in facilitating molecular marker analysis, which allows prognostication as well as individualization of chemotherapy 32. In 2002 the Trucut biopsy (TCB) needle was introduced 33. Unfortunately the Trucut was not widely embraced due to its relatively lower diagnostic yield, high failure rate particularly in the duodenum and slightly higher complication rate 34, 35. However, recent introduction of the core histology needle (CHN) (Cook Medical, Winston-Salem, NC, USA) with an integrated reverse side bevel has generated significant interest and data from studies performed by various groups. Most recently, a multicenter prospective RCT published in abstract form showed FNB having higher diagnostic yield than FNA particularly for non-pancreatic lesions (86% vs 55%, P = 0.02) 36. Katanuma et al. 37 recently published a very nice article which quantified the amount of suction generated through a variety of needle sizes by various suction techniques. This study revealed the suction pressure at the end of a straight or CHN needle to be essentially identical, therefore, the differences in specimen and diagnostic yield, is likely to be related to the structural differences. Currently, the CHN is available in 19 g, 22 g and 25 g. The idea of the CHN was to facilitate a combination of FNA and FNB (FNAB) which would then result in acquisition of a large volume of tissue (Fig. 2) with intact cores for histological analysis. This is advantageous when ancillary testing is required, with previously failed FNA, and when rapid onsite evaluation (ROSE) is unavailable or simply to augment cytology to further increase the overall diagnostic yield 38. Cores are more easily recognized on gross examination of the glass slide compared to clumps of cells, as they have a whitish appearance and maintain some height or bulk on visual inspection (Fig. 2). The recognition of core tissue without microscopy, may lead to higher specimen adequacy and subsequent improved diagnostic yield, especially in the absence of ROSE 39. The use of the 19 g CHN (Fig. 3) was first published in 2011 and has been shown to be technically feasible (98–100%) and produced adequate histological specimen (89.5–94.2%) 40, 41. The 22 g CHN (Fig. 3), when used for intra-abdominal solid mass, has high yield both cytologically and histologically (Fig. 4) with 76% having useful core specimen for pathological diagnosis 42. Performance on pancreatic mass lesions showed a one pass result of 98% feasibility and 88.5% adequate core for histological analysis 43. A randomized trial comparing the 22 CHN vs 22 straight was published in abstract form where 97 patients with mixed lesions were randomized to either needle, with all specimens submitted in cell block only, resulted in a diagnostic yield of 92% (CHN) vs 77% (P = 0.03) 39. In studies comparing number of passes, the 22 g CHN required less in a cohort of mixed pancreatic lesion and lymphadenopathy (1.2 vs 2.5, P ≤ 0.001) 44 and non-pancreatic lesions (2.11 vs 2.94, P = 0.03) 45. However, no difference was seen when restricted to pancreatic lesions 46. One small study compared performance in gastrointestinal subepithelial lesions where the 22 g CHN yielded significantly higher macroscopic (92% vs 30%, P = 0.006) and histological cores (75% vs 20%, P = 0.01) in fewer median number of passes (2 vs 4, P = 0.025) 47. The 25 g CHN has been reported by Iwashita et al. to have a high yield sensitivity of 83%, 91%, and 96% on pass 1, 2, 3, respectively 48. Visible core (Fig. 5) was seen macroscopically in 92% of cases, however histological core only in 32%. All procedures were uncomplicated. An abstract presented at DDW 2014 showed a high cellular and diagnostic yield using the 25 g CHN in absence of ROSE 49. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of EUS-FNAB in which sampling adequacy was assessed by the endosonographer were 93.3%, 100%, 100%, 93.3%, and 96.6% (for cytologic diagnosis) and 92.3%, 100%, 100%, 92.9%, and 96.2% (for histology), respectively. When comparing the 25 g straight to the 25 g CHN, there is little evidence in the literature. One study showed the tissue acquisition rate for histological analysis was 90.6% (CHN) vs 79.6% (straight) (P = 0.025) with definitive histological diagnosis achieved by the 25 g CHN in 81.1% (CHN) vs 69.4% (straight) (P = 0.048) 50. Although the 19 g straight needle was not specifically designed to improve histological yield, its ability to acquire histological material has improved over time. Multiple studies published in 2012 demonstrated high diagnostic yield and accuracy to be in the 90% or higher for pancreatic mass 51, lymphadenopathy 19 or submucosal lesions 12, 20. To date, there has not been a trial comparing the 19 g CHN to the 19 g straight needle. A 22 g side port needle (Olympus, Tokyo, Japan) has been tested in a pilot study 52 and a small study of 30 patients with mixed lesions. The results have been impressive with a diagnostic accuracy of 96.7% after only 1.7 (mean) passes without any complications 53. However, further studies are warranted to confirm this pilot data. An 18 g Cryoprobe (Erbe, Tübingen, Germany) had been trialed in non-survival animal and cadaveric pancreas showing a comparable result to the Trucut and better than 19 g straight needle in terms of tissue specimen quality and size and shorter bleeding time 54. This probe uses carbon dioxide induced cooling to minus 35 degrees, which freezes the tissue without causing apparent histological tissue damage. Specimen extraction is possible due to cryoadhesive effect and can be performed with or without an outer sheath through the scope. There are no safety data regarding this novel modality of tissue acquisition. When histological tissue is required, the needle of choice should be the 19 g CHN, 22 g CHN or 19 g straight needle. The benefit of suction is controversial and an established standard is lacking. Varying degrees of suction can be applied from slow pull of the stylet from the needle, which in turn generates microsuction, to 5 or 10 ml of syringe suction, to continuously high (30 ml) suction facilitated by the Alliance II inflation system (Boston Scientific, Natick, MA, USA) 55 termed EUS-FNTA. By using suction, a higher volume of aspirate can be acquired, which may or may not be useful depending on the situation. According to the Katanuma study 37 maximal suction power (when attached to a suction syringe) of a 19 g needle is approximately two to three times and seven to eight times more than the 22 g and 25 g needle, respectively. This is the likely mechanism for increased specimen yield when using the largest needle primarily for histology. Certainly suction can be attempted in situations where cellularity is inadequate without suction or in fibrotic lesions 56, whereas in more vascular lesions such as lymph nodes, suction may cause more bleeding and therefore negatively affect the specimen quality 57-59. Two RCTs found conflicting results in FNA of pancreatic mass lesions where one found suction resulted in higher diagnostic samples and cellularity; however, specimen bloodiness was increased 60 and the other found no difference in diagnostic yield, cellularity, or specimen bloodiness, irrespective of whether or not suction was used 61. A trial comparing the slow pull with the suction technique showed better sensitivity and less blood contamination when using a 25 g FNA needle but no significant difference when using the bigger 22 g needle for pancreatic mass 62. The Katanuma study 37 revealed that the suction pressure of a 25 g needle without suction vs slow pull was very similar and possibly below the level of tissue pressure, therefore reducing the likelihood of a bloody aspirate. In animal studies, when using the CHN, ironically slow pull acquired larger tissue volume than suction 63. In a study of 50 patients, a very high first pass diagnostic accuracy of 86% was achievable by combining the 25 g CHN with the slow pull technique 48; however, no comparison with other techniques was made. Two studies investigating the high suction FNTA method using a 22 g needle showed a diagnostic accuracy of approximately 77% 55, 59 with one pass. Given suction did not clearly show a benefit, this led investigators to modify the conduit, which transmits the negative pressure within the needle. Given that fluid is less compressible than air, greater negative pressure therefore can be transmitted to the end of the needle. The results of a computer generated model was recently published in abstract form, which showed a 70% theoretical increase in total volume of tissue aspirated with fluid primed needle vs dry suction 64. Clinically, using normal saline primed needle and 10 ml of suction during FNA yielded superior cellularity (1.83 vs 1.33, P = 0.001) and diagnostic rate (85.5% vs 74.4%, P = 0.0001) compared to dry suction technique when using a 22 g needle 65. However, when using heparin instead of normal saline in a group of 60 patients, there was no difference in diagnostic adequacy, number of passes or specimen bloodiness 66. Another interesting study published in abstract form compared three different suction techniques 67, namely, dry, wet and hybrid. Dry suction was described as the conventional method of stylet removal after puncture, application of continuous 10 ml of syringe suction followed by 12 to and fro movements. The wet technique involved priming the needle with normal saline by attaching a 10 ml syringe prefilled with 3 ml of normal saline. FNA was performed with three to and fro movements first then applying 10 ml of suction. The to and fro then suction maneuver was repeated four times to total 12 to and fro movements. The hybrid technique involved the same needle preparation as the wet technique, but with constant 10 ml of suction. The difference in specimen adequacy was 87% (wet), 87% (hybrid) and 67% (dry) and diagnostic yield was 100% (wet), 92% (hybrid) and 90% (dry) which was not statistically significant. This may be due to an under powered study as only a total of 45 FNA passes were performed on 15 solid lesions. The theory of using stylet is to minimize contaminants from entering and obstructing the lumen of the FNA needle, which result from the initial puncture. However, a large retrospective study involving over 3000 patients 68 and a prospective comparison study where the same lesion was sampled with and without stylet showed no difference in diagnostic yield 69. Furthermore, two RCTs also showed no difference in diagnostic yield with or without the use of a stylet 70, 71. However, these studies examined predominantly 22 and 25 g needle, not the 19 g. Therefore, advocacy for stylet usage is most likely dependent on personal anecdotes. The 22 g needle did not confer any advantage over the smaller 25 g when no stylet was used 72. The stylet appears to be the best method of pushing the cells out of the needle, therefore, having the stylet remain in the needle, as with the “slow-pull” technique, reduces the stylet handling, as the stylet never leaves the needle. Diagnosis of typical lesions are usually straightforward, however, this can be augmented with the fanning technique described by Bang et al. 73 in a randomized trial showed almost 30% increase in diagnostic yield in the first pass. This technique makes logical sense in that all areas of the lesion are systematically sampled instead of targeting the same tract repeatedly or randomly. However, in very large lesions often there is central necrosis. Aiming for the periphery in this instance may avoid the necrotic areas hence increasing diagnostic yield 56; however, this principle does not necessarily apply to lymph node FNA 57. Number of passes and specimen interpretation are intimately related. The optimal number of passes varies depending on the target characteristics and the availability of ROSE. For example, a well differentiated adenocarcinoma may be more difficult to delineate at EUS or be called a definite positive at ROSE, therefore resulting in a higher number of passes 70. It has been reported that three passed for malignant lymph nodes or liver lesions have a very high yield 57, 74. However, in absence of ROSE, the optimal number may be increased for lymph nodes and solid lesions to five and five to seven, respectively 74, 75. Submucosal lesions can be equally difficult to diagnose and performing up to five passes is recommended 13 although one study described plateauing yield after four passes 58. Regarding solid pancreatic lesions, four passes with a 25 g needle may be sufficient with or without ROSE as the yield had been shown to be similar 76. Certainly one of the key factors in minimizing unnecessary passes thereby reducing potential complications, improving efficiency and yet maintaining a high diagnostic yield is the availability of ROSE in most cases. Although ROSE can dramatically improve yield while reducing number of passes performed 77-79, the absolute benefit will vary depending on the pre-existing diagnostic yield by the endosonographer without ROSE. These findings have also been supported by a recent systematic review and meta-analysis 80. However, if the cytopathologist is not available, endosonographer initiated ROSE will also significantly improve diagnostic adequacy 81. The impact of specimen extraction or expression from the FNA needle is poorly studied. Typically the stylet is reinserted in order to push out the acquired specimen. Theoretically air or fluid flush is more complete as there is residual space between the stylet and the needle lumen. A prospective comparative trial investigated specimen expression with conventional stylet insertion vs air flush showed specimens are slightly less bloody with air flush without any impact on diagnostic yield 60. Stylet may still be needed if the needle is obstructed by clot; however, if expelled in a timely matter, air flush is usually adequate 69. Blood and clot should be separated from the specimen slide if possible, as this optimizes the cytological interpretation. For histologic core samples, blood usually does not interfere with histologic interpretation; however, a large amount of blood clots may potentially produce problem blocks. Another consideration regarding specimen preparation is how much material should be committed to preparing a cytology slide as opposed to directly placing into a formalin bottle. The advantage of cytology is the potential to provide an immediate preliminary diagnosis; however, there are several disadvantages including longer procedure time, longer time to official diagnosis, utilizing more resources per case and lower quality of histology due to loss of micro cores to the cytology slide 32. One alternative option is to use solutions such as Cytolyt in which the entire specimen is sent to pathology, and both cytology slides and cell block can be prepared from the sample. A DDW abstract comparing yield between Cytolyt and formalin did not a significant difference if adequate cytology is not available, then one all material into cell block directly without any The use of stylet may be however, the technique when using the 25 g needle appears to be advantageous. suction is an interesting and further of the fanning technique and of necrotic will increase diagnostic ROSE by cytopathologist or endosonographer will improve yield therefore reducing the number of passes required. One should more tissue to cell block if histological diagnosis is a cytologic diagnosis is by ROSE, to the larger histology needles with of the entire specimen for histology. Given the ability to core tissue by EUS the potential for liver biopsy has been studies used the 19 g with success adequate specimen quality been as complete and of specimen specimen from were frequently inadequate only having a patients in their This is particularly as patients are to be more difficult to biopsy as the specimen is to and shorter specimen yield In the first case of autoimmune diagnosed on liver biopsy with a 19 g CHN was published and around that time, et al. published their with a 19 g straight needle where 22 patients were by needle passes a median and of and and a very high diagnostic yield of without any complications or significant years et al. also used a 19 straight needle in 10 patients had liver with two to be The was and was after three passes without complications or significant accuracy was 100% and this only an of to the EUS procedure. data were published in abstract form showing results in over patients using the 19 g straight needle with median and of 17 and when both and were However, characteristics and FNA technique was not on initial pilot studies, EUS guided liver biopsy appears to be safe and well by patients. However, further studies with on patients are needed. There are several factors that can affect the performance of EUS One is the choice of needle and using the available evidence as a the the (Fig. as a to the most needle. factors including fanning technique, slow pull when using the 25 g needle, no stylet or suction for the 22 g and using the core histology needle when ancillary testing is required are also of needle designs and technique, this will no lead to higher specimen yield as well as the of EUS guided FNA or is a to technique during ultrasound guided fine needle aspiration (EUS pull Endoscopic ultrasound guided liver biopsy specimen The is not for the or of any by the than should be to the for the

Endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) with rapid on-site evaluation (ROSE) has been reported to provide a more accurate diagnosis than EUS-FNA without such evaluation. However, even endosonographers can evaluate ROSE regarding sample adequacy. The aim of this study was to evaluate the diagnostic accuracy of EUS-FNA with ROSE by endosonographers compared to ROSE by cytopathologists in patients with solid pancreatic masses. Between September 2001 and October 2005, of the 73 EUS-FNA procedures with the final diagnoses, 38 procedures after the introduction of ROSE by endosonographers (September 2001-September 2003, period 1), and 35 procedures after the introduction of ROSE by cytopathologists (October 2003-October 2005, period 2) were included. The specimens were stained with Diff-Quik stain and assessed. When the on-site assessors (endosonographers or cytopathologists) indicated that the amounts of cell samples were adequate, the procedure was stopped. Results are presented with 95% confidence limits. The average numbers of needle passes were 4.0 +/- 1.6 and 3.4 +/- 1.5 in periods 1 and 2, respectively (P = 0.06). The specimen collection rates were 97.4 and 97.1% in periods 1 and 2, respectively (P = 0.51). Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for malignancy and benign were 92.9, 100, 100, 83.3, and 94.7%, respectively, in period 1, and 93.1, 100, 100, 75.0, and 94.3%, respectively, in period 2 (P = 0.97, P = 1.0, P = 1.0, P = 0.65, P = 0.93, respectively). No complications were seen. For accurate diagnosis, ROSE should be performed during EUS-FNA by the endosonographer, if no cytopathologist is available.

Practical Guidance on Establishing a Molecular Testing Pathway for Alterations in Homologous Recombination Repair Genes in Clinical Practice for Patients with Metastatic Prostate Cancer