Endoscopic ultrasound (EUS) is an indispensable tool for tissue acquisition in patients with gastrointestinal tumors. While fine-needle aspiration (FNA) has been routinely carried out for establishing tissue diagnosis, the emerging concept of tailoring chemotherapeutic agents based on molecular markers has increased the demand for core tissue procurement by means of EUS-guided fine-needle biopsy (EUS-FNB). In addition, FNB may offset the limitations of FNA wherein the diagnostic sensitivity is incumbent on the availability of an onsite cytopathologist. Given the increasing number of procedures being done, developing a unit-specific algorithmic approach for needle selection isimportant to improve the procedural efficiency and utilization of resources. Finally, the best outcomes can be attained only by practicing evidence-based techniques, procuring adequate quantity of sample for ancillary studies and processing the specimens appropriately.
The objectives of the present review are to summarize three key aspects in tissue acquisition that are important for an endosonographer: (i) determine the type of tissue to be procured based on clinical need and availability of resources; (ii) identify factors that impact the efficiency of an endoscopic ultrasound (EUS) service; and (iii) practice evidence-based techniques to improve the clinical outcomes of the procedure.
Tissue Procurement: Aspirate versus Biopsy
Tissue procured at EUS can be either a fine-needle aspirate (FNA) or a fine-needle biopsy (FNB). The choice of tissue should be based on: (i) clinical need; and (ii) availability of resources. In most patients, an FNA is sufficient to establish a diagnosis. The technique is quick, safe, accurate, inexpensive, and the interpretation is reliable when carried out by a trained cytopathologist. However, the sensitivity of EUS-FNA is superior only when the specimen is assessed onsite for diagnostic adequacy.[2-4] This is because a cytological ‘aspirate’, unlike a ‘biopsy’ specimen, is quantitatively less and could be insufficient for diagnostic evaluation. Therefore, endosonographers prefer onsite assessment for diagnostic adequacy. Processing an FNA sample in the procedure room is technically easy. When a microscope and the requisite stains are available, the slides can be prepared within a few minutes and a preliminary diagnosis rendered real-time during the procedure. In addition, this approach facilitates a series of specimens to be collected, which allow additional testing such as flow cytometry, immunohistochemistry or microbiological analysis to be accomplished during the same procedure. Also, a very significant but unquantifiable advantage of onsite assessment is that patient care can be expedited. This includes evaluation by subspecialists or sometimes even preoperative assessments to take place the same day. However, a major limitation to this approach is that most institutions do not have skilled individuals to render onsite assessment as the majority of the workload consists of surgical pathology and not cytopathology.
A FNB specimen contains core tissue with better preservation of cellular architecture than an aspirate. Therefore, in general, a FNB specimen has greater diagnostic accuracy and provides more tissue for ancillary testing than a typical FNA sample. While a specialized needle may oftentimes be required to procure core tissue, the choice of needle does not exclusively determine the definition of ‘core’ as this can be yielded intentionally or by chance using any needle. A core tissue may be specifically requested by pathologists to establish a definitive diagnosis in challenging cases, such as in well-differentiated pancreatic cancer or autoimmune pancreatitis, when FNA is inconclusive.[7, 8] Other indications for FNB procurement include identification of molecular markers that are specific for neoplasms such as pancreatic neuroendocrine tumors and metastatic lung or breast cancer so that they can be specifically targeted with therapeutic agents.[9, 10] Unfortunately, like FNA, it is possible to encounter sampling errors with FNB. Therefore, rather than collecting the core tissue in formalin and sending it to the pathology lab for processing and interpretation the next day, endosonographers emphasize onsite assessment for adequacy. However, assessing the adequacy of a FNB sample can be more challenging. Problems are related to the nature of cell transfer from the FNB to the glass slide, which is accomplished by making multiple touches of the tissue core to the slide (touch prep) or rolling the core tissue over the slide. Compared to an FNA specimen, these maneuvers typically transfer significantly fewer cells and can cause cellular distortion and promote air drying, thereby negatively impacting on the adequacy of the assessment. Furthermore, the physical manipulation of the core tissue and delay in placing the small core into a fixative medium can sometimes lead to tissue disruption and poor preservation that can cause difficulty in histology interpretation.
There is no one solution or answer to all our unmet needs in tissue acquisition. Therefore, at a given institution, endosonographers and pathologists, based on the availability of resources, must collectively devise a strategy that yields maximum accuracy under most circumstances. For institutions that have access to onsite cytopathology support, an efficient option would be to carry out an FNA and assess diagnostic adequacy or establish a preliminary diagnosis during the procedure itself. Several studies have shown that the diagnostic accuracy of EUS-FNA for evaluation of solid pancreatic masses and lymph nodes is >90% and 95%, respectively.[13, 14] Based on FNA findings, a FNB may be carried out only if additional tissue is required for ancillary studies or if a core biopsy is specifically requested, such as for assessment of molecular markers. In lieu of a core biopsy, multiple FNA passes can be done in the standard fashion and the specimen collected in CytoLyt (methanol-water solution), CytoRich Red, Roswell Park Memorial Institute media (RPMI), or Hanks Balanced Salt Solution (HBSS) for creation of a ‘cell block’. A cell block is created by centrifuging the small FNA fragments and combining them with a clotting agent, whether plasma and thrombin, albumin, or histogel. The clot is then paraffin-embedded, sectioned and stained with hematoxylin and eosin to facilitate histopathological evaluation. Studies have shown that a cell block is a reliable technique for histological evaluation of specimens, carrying out immunostaining, and for assessment of molecular markers and proteonomics.[15, 16] Therefore, most indications for a core biopsy can be met by carrying out adequate FNA passes for cell block preparation.
For institutions that do not have access to onsite cytopathology support, options include carrying out at least three FNB passes using a core biopsy needle or a 19-G needle, or carrying out multiple FNA passes for the preparation of a cell block. Two prospective studies on EUS-guided biopsy of pancreatic masses have shown that by using a dedicated core biopsy needle or a 19-G needle, a satisfactory histological assessment can be made in >90% of patients.[17, 18] However, the number of dedicated passes that one has to carry out in order to obtain an adequate cell block is unclear. Indirect evidence suggests that for the evaluation of pancreatic masses, a single pass with a 19-G needle and a 25-G needle yields a cell block in 85% and 33% of patients, respectively. In a recent study on FNA of pancreatic masses in 20 patients, carrying out four dedicated passes with a 25-G needle established a definitive diagnosis in 100% of patients. In the opinion of these authors, at least three dedicated passes must be carried out with a 19-G needle or a core biopsy needle and four or more passes must be done when using a 22- or a 25-G needle to obtain a diagnostic cell block. Randomized comparative trials are required to answer this important question.
Factors that Impact the Efficiency of an EUS Service
There are three issues that are critical to the efficiency of an EUS service: the availability of onsite cytopathology support, choosing appropriate needles to sample a lesion, and carrying out FNA with maneuvers that yield a quick but accurate diagnosis.
On-site cytopathology support
Studies have shown that the presence of a cytopathologist during EUS-FNA improves the diagnostic yield, decreases the number of inadequate or unsatisfactory samples and limits the number of passes required to establish a diagnosis[2-4, 21] (Table 1). For institutions limited by resources, carrying out FNB or collecting adequate specimens for cell block preparation is important. In a recent study of 138 patients who underwent EUS-FNA of solid pancreatic mass lesions, two endosonographers carried out cytopathology procurement and onsite assessment for diagnostic adequacy themselves. Compared to the time period when all specimens (n = 53) were evaluated off-site, after being proctored in cytopathology, the same endosonographers carried out onsite assessment on all subsequent pancreatic mass FNA (n = 85) and found that the rates of diagnostic accuracy improved by 22% and the rate of inconclusive diagnosis decreased by 18%.
Table 1. Studies evaluating the role of onsite cytopathology support for EUS-guided FNA procedures
Diagnostic yield OP vs no OP
Indeterminate/Inadequate samples OP vs no OP
Unsatisfactory OP vs no OP
aEndoscopists interpreted specimens themselves after being proctored in cytopathology.
EUS, endoscopic ultrasound; FNA, fine-needle aspiration; NA, data not available; OP, onsite cytopathology.
EUS-FNA can be carried out using a 25-, 22- or a 19-gauge needle. The choice of needle should be based on the: (i) ability to obtain an adequate amount of cellular material to establish a diagnosis; (ii) flexibility to access a lesion successfully; (iii) low risk of complications; and (iv) option to acquire core tissue when required. Five randomized trials and one meta-analysis[19, 23-27] have attempted to identify the optimal needle for carrying out FNA (Table 2). Of the three studies that compared the 22- and 25-G needles,[23-25] while there was no significant difference in diagnostic accuracy, there was a trend towards better performance with the 25-G needle for FNA of pancreatic head/uncinate masses. In one study that compared the 19- and 22-G needles, better diagnostic accuracy and cellular material were acquired using the 19-G needle. However, there was a significantly higher technical failure rate with the 19-G needle for FNA of pancreatic head masses. A recent multicenter randomized trial compared the 19-G and 25-G needles for EUS-guided FNA of solid pancreatic mass lesions. While there was no significant difference in diagnostic accuracy or technical failure between both needles, the 19-G needle yielded a histological core tissue in a significantly larger cohort of patients, 85 vs 33%. In this study, the investigators used a 19-G needle made of nitinol (Flex 19; Boston Scientific, Natick, MA, USA) with enhanced flexibility properties, to carry out transduodenal FNA. In a meta-analysis that compared the 25-G and 22-G needles for EUS-FNA of solid pancreatic mass lesions, the diagnostic sensitivity of the 25-G needle was significantly better than that of the 22-G needle.
Table 2. Randomized trials and meta-analysis comparing different needles for EUS-FNA of solid mass lesions
Diagnostic accuracy/pooled sensitivity in meta-analysis
EUS, endoscopic ultrasound; FNA, fine-needle aspiration; NA, data not available; NS, not significant; RT, randomized trial.
85 (95% CI: 82–88%) vs 93% (95% CI: 91–96%), P = 0.0003
There are no data evaluating the different types of needles for cyst aspiration. In terms of efficiency, a 19-G needle is most suited for carrying out cyst aspirations. However, if the cyst is small (<2 cm), a 22-G needle may be easier to use for aspirating the cyst contents.
A dedicated 19-G Trucut (Cook Endoscopy, Winston-Salem, NC, USA) biopsy needle (EUS-TNB) was developed nearly a decade ago to procure core tissue for histological analysis. However, the rigidity induced by its 19-G caliber and the mechanical friction of the firing mechanism produced by the torqued echoendoscope limited its use for carrying out transduodenal biopsies. To overcome this limitation, a new 19-G FNB device was recently developed with ProCore (ProCore; Cook Endoscopy) reverse bevel technology to enable the acquisition of core specimens. In a recent study from Europe, histological samples were obtained successfully with this ProCore needle in a majority of patients with a diagnostic accuracy of >90%. However, some technical difficulties were encountered when carrying out transduodenal passes. The same FNB device is also available in a 22-G and 25-G platform to facilitate easy transduodenal sampling.[30, 31] In a prospective study of 61 patients, EUS-FNA of solid pancreatic masses done using a 22-G needle established a histological diagnosis in 88.5% of patients. Although a cytological diagnosis was established in 96% of 50 patients undergoing EUS-guided sampling of solid pancreatic masses using a 25-G ProCore needle, histological core tissue was procured in only 32% of patients. The role of the standard 19-G FNA needle for yielding histological samples was assessed prospectively in a recent study. Of the 120 patients who underwent EUS-guided tissue acquisition, the procedure was technically successful in 119 patients (98.9%) and adequate histological sample was obtained in 116 (97.5%). A major limitation of the study was that patients with pancreatic head or uncinate masses were excluded. As the standard 19-G needle is too stiff to navigate the transduodenal route, a flexible 19-G needle made of nitinol has been recently introduced. In a pilot study of 50 patients, which included several patients that underwent EUS-FNA via the transduodenal route, histological specimen was procured in 95% of patients and no technical failure was encountered in any of the patients.
The safety profile of EUS-FNA procedures is excellent. In all of the studies that have compared the three needle types,[19, 23-26] there was no significant difference in rates of complications between the cohorts. In a meta-analysis that compared the 19-G versus 22- and 25-G needles for EUS-FNA of pancreatic masses, there was no significant difference in rates of pancreatitis, bleeding, infection, perforation or abdominal pain between both cohorts. However, experience with the use of the 19-G needle for tissue acquisition is limited.
Data on resource utilization during EUS-FNA procedures are limited. Recently, a prospective, two-phase, algorithm-based study evaluated resource use during EUS-FNA procedures. In phase I, 548 EUS-FNA procedures were carried out using a 22- or 25-G needle interchangeably and the technical failure rate (need for use of more than one needle for FNA of an individual lesion as a result of technical dysfunction) was noted to be 11.5%. Based on these observations, an algorithm was proposed by which all transduodenal FNA were carried out by using a 25-G needle and others with a 22-G needle. This algorithm was tested prospectively in phase II on 500 patients. The technical failure rate was significantly less in phase II compared to phase I (1.8% vs 10.9%; P < 0.001). Although there was no difference in diagnostic adequacy (97.1% vs 98.4%; P = 0.191) or complications (0.4% vs 0.2%; P = 1.0) between both phases, the average cost per case was significantly less in phase II by $11.
The center of a cancerous mass tends to be more necrotic than the periphery and hence yields non-viable material when sampled. Therefore, to overcome this limitation, two studies have suggested that aspirating a lesion at its periphery or in multiple areas may improve the diagnostic yield.[36, 37] In the ‘fanning’ technique of FNA, at individual passes, the needle is positioned at four different areas within a mass and then moved back and forth four times in each area to procure tissue. We term this the ‘4 × 4’ rule. Aspiration is usually initiated at the left margin of the tumor mass and then ‘fanned’ until the right margin of the tumor is sampled. The trajectory of the needle can be altered using either the ‘up/down’ endoscope dial or the elevator. In a recent randomized trial of 54 patients with solid pancreatic mass lesions, the fanning technique established a significantly higher first-pass diagnosis in 85.7% of patients compared to only 57.7% with the standard technique. Also, the median number of passes required to establish a definitive diagnosis was significantly less with the fanning technique. While the fanning technique seems to have a distinctive advantage when sampling a primary tumor mass, its effectiveness was not demonstrated in lymph-node sampling. Although recent advances in contrast-enhanced imaging and elastography show promise in differentiating a malignant from a benign tumor, its ability to ‘guide’ the needle for sampling an area of viable tumor within a necrotic mass is unknown.
Although manufacturer guidelines must be followed when using specially designed biopsy needles, there are no objective data on the best technique to adopt when using a 19-G needle. In our experience, when using a 19-G needle, to minimize bloodiness, one must not use suction or a stylet. Although the needle must be carefully ‘fanned’ within the mass, repeated ‘jabbing’ at one area should be avoided. Rather, at each location within the mass, the needle must be moved back and forth only two to three times (4 × 2 rule). It is usually not necessary to carry out more than three FNB passes in a lesion as repeated biopsies are more likely to yield blood clots.
The presence of an onsite cytopathology team improves the yield and decreases the number of passes required to establish a diagnosis. When onsite cytopathology support is not available, carrying out at least four to five FNA passes for cell block or carrying out three FNB for histopathological analysis is important. Although data are limited, learning the basics of cytopathology may help endosonographers to independently assess for diagnostic sufficiency and thereby improve the overall procedural outcomes. Incorporating an algorithmic approach to FNA and FNB not only improves the technical outcomes but also optimizes resource utilization (Fig. 1). The preponderance of data suggests that thinner and flexible needles are better suited to sampling pancreatic masses, particularly those accessed via the transduodenal route. If histopathology is desired, a specially designed biopsy needle or a 19-G needle can be used to procure core tissue. Finally, the ‘fanning’ technique of FNA improves procedural efficiency by establishing a first-pass diagnosis in a majority of patients.
Recognition of Factors that Improve Technical Outcomes
It is important to recognize that there are several factors that determine the outcomes of EUS-FNA: location of the lesion, use of suction to aspirate tissue, use of a stylet in the needle assembly, experience level of the endosonographer; and adopting appropriate safety measures.
In general, whereas transesophageal and transgastric lesions are the easiest to sample, transduodenal lesions are more difficult. In addition to the fact that the anatomical landmarks of the pancreatic head are more complex, the torqued echoendoscope makes transduodenal sampling difficult by precluding easy deployment of the FNA needle. As emphasized earlier, choosing a smaller 25-G or flexible needle makes the puncture technically easier. It is also important to know when to terminate the procedure. If the needle is in the correct location and an onsite diagnosis cannot be established within seven to eight passes in a pancreatic mass or after five passes in a lymph node, one reaches a point of ‘diminishing return’. If the clinical suspicion remains high, carrying out the procedure on a different day or obtaining a surgical biopsy is likely to yield a more successful outcome than persisting with more FNA passes.
Use of suction
Although the use of suction increases the quantity of the sample, it makes the specimen bloodier and diminishes the quality of the aspirate. One randomized trial on FNA of lymph nodes revealed that the use of suction yielded a bloodier specimen but without any difference in diagnostic accuracy. In another randomized trial on pancreatic masses, there was no difference in diagnostic yield, cellularity or specimen bloodiness, irrespective of whether or not suction was used. In routine clinical practice, we do not use suction in any patient unless the lesion is cystic or if the aspirate is a dry tap. In several studies on FNA of solid pancreatic masses, we were able to establish a diagnosis without the use of suction in >90% of patients.[17, 19, 35]
Use of a stylet
The intent of a stylet is to prevent the tip of the needle from being clogged by a plug of digestive wall tissue when a lesion is sampled. A total of four randomized trials have evaluated the role of a stylet.[43-46] In three studies, the use of a stylet did not improve the diagnostic yield for malignancy.[43-45] In one study of 81 patients with pancreatic masses, the diagnostic yield, sensitivity, accuracy, cellularity and bloodiness were significantly better with the use of a stylet. Also, flushing the needle with air rather than using a stylet to express the specimen resulted in less bloody specimens. In our opinion, the routine use of a stylet is cumbersome and increases procedural duration. We also believe that it contributes to increased specimen bloodiness without any measurable difference in diagnostic yield. However, we do not discard the stylet but rather use it for controlled expression of the aspirate on to individual slides.
Experience level of the endosonographer
Although some professional societies recommend a minimum of 25 pancreatic and 25 non-pancreatic FNA procedures to attain competency, two studies have examined the learning curve associated with EUS-FNA of pancreatic masses.[48, 49] In the first study of 57 patients who underwent EUS-FNA by a self-taught endosonographer, the diagnostic sensitivities for malignancy from the first to the last 10 quintile were 30%, 40%, 70%, 90%, and 80%, respectively. In another study of 300 consecutive patients who underwent EUS-FNA by a trained endosonographer, the median number of passes required to establish a diagnosis decreased significantly with operator experience but without any difference in diagnostic accuracy. We certainly believe that there is a learning curve for carrying out EUS-FNA of pancreatic masses. While the diagnostic accuracy may plateau over time, the procedural expertise continues to improve with experience.
The majority of complications in EUS-FNA are encountered during aspiration of cystic lesions. Therefore, prophylactic antibiotics must be given and continued for up to 48 h for EUS-FNA of cystic lesions. When carrying out a FNA, a change in echogenicity denotes intracystic bleeding. Clinically significant intracystic hemorrhage is rare and most bleeding usually resolves spontaneously. Whereas low-dose aspirin can be safely continued, clopidogrel must be discontinued 7 days prior to EUS-FNA. Low-molecular weight heparin must be discontinued 12–24 h prior to FNA and unfractionated heparin must be stopped 6 h prior to procedure. It is recommended that warfarin be discontinued 5 days prior to EUS-FNA in low-risk patients or be bridged with heparin in patients at high risk for thrombotic events.
Currently available evidence suggests that suction should not be used routinely for all procedures unless the aspirate is a dry tap or if a cyst needs to be aspirated. Likewise, the routine use of a stylet can be cumbersome and can increase specimen bloodiness but without a significant impact on diagnostic yield. It is important to be aware that carrying out ‘innumerable’ passes on a single lesion can be meaningless after a certain point. Repeating the procedure on a different day is more likely to yield a successful outcome. Finally, like any other endoscopic procedure, with time and experience, one gets better at carrying out FNA.
In the present review, we have attempted to provide a perspective by reviewing several key issues. EUS-guided tissue acquisition is a multi-step procedure. Knowing the procedural indication, procuring the correct type of sample by identifying the needs of the treating physician, selecting the appropriate needle based on the location of the lesion and type of tissue to be procured, sampling the lesion using the best evidence-based techniques, procuring additional tissue for ancillary studies, processing the specimens appropriately and closely collaborating with cytopathologists are all important for good technical and clinical outcomes.
Conflict of Interests
Authors declare no conflict of interests for this article.