Endoscopy 2018; 50(05): 466-468
DOI: 10.1055/s-0043-125211
© Georg Thieme Verlag KG Stuttgart · New York

Endoscopic ultrasound-guided fine-needle aspiration of solid pancreatic lesions: striving for perfection

Referring to Saxena P et al. p. 497–504
John Gásdal Karstensen
1  Department of Gastrointestinal Surgery, Slagelse Hospital, Slagelse, Denmark
Peter Vilmann
2  Gastro Unit, Division of Endoscopy, Copenhagen University Hospital, Herlev, Denmark
› Author Affiliations
Further Information

Publication History

Publication Date:
25 April 2018 (online)

Endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) is the procedure of choice whenever a pathological diagnosis is needed from a solid pancreatic lesion [1]. However, despite nearly three decades of intensive efforts to increase the diagnostic yield, the sensitivity of EUS-FNA remains suboptimal. Meta-analyses of randomized controlled trials (RCTs) evaluating EUS-FNA in solid pancreatic masses have shown that the sensitivity is in the range of 85 % – 93 % [2] [3]. Extensive efforts have been made to further optimize the EUS-FNA procedure. Although rapid onsite cytology (ROSE) has consistently been shown to increase the sensitivity of EUS-FNA, an updated meta-analysis has recently questioned whether ROSE is associated with an increase in diagnostic yield, rate of adequacy, pooled sensitivity, and specificity [4]. Different sizes and designs of needles have been tested in several well-conducted RCTs without improving the sensitivity [5] [6]. Similarly, different suction techniques and use of needles with or without a stylet have undergone RCTs without really pushing the diagnostic boundaries of the EUS-FNA procedure [7] [8].

“Further optimization of technical procedural issues with EUS-FNA of solid pancreatic lesions seems to have reached a plateau.”

In this issue of Endoscopy, Saxena et al. present an RCT evaluating the effect of the stylet slow-pull technique vs. a standard suction technique with a 22-gauge needle (Expect; Boston Scientific Corp., Marlborough, Massachusetts, USA) for diagnosis of solid pancreatic masses [9]. While one drawback of the suction technique is possible contamination of the specimen with blood, the authors hypothesize that the slow withdrawal of the stylet in the slow-pull technique applies minimum negative pressure in the needle, which might optimize the quality of the specimens, and hence, increase the sensitivity of the procedure. A total of 121 patients were enrolled in the study and subsequently randomized to either the stylet slow-pull technique or the suction technique. The patients were recruited from two high-volume US centers, and FNA passes were conducted with ROSE assistance in all cases. There was no difference between the two groups in terms of diagnostic sensitivity (primary end point), first-pass diagnostic rate, or median number of passes to diagnosis. These results extend the general impression that technical procedural changes may not increase the diagnostic yield of EUS-FNA, and that the sensitivity may have reached an upper limit for solid pancreatic lesions. Obviously, attempts to push the boundary further are most welcome; however, given the virtually perfect specificity of the procedure, we believe that the sensitivity of EUS-FNA for solid pancreatic masses has reached a plateau at this level.

Despite negative results of the study, we want to congratulate Saxena et al. for a beautifully conducted trial. Only RCTs can truly evaluate specific interventions such as the stylet slow-pull technique, and it is because of high-quality studies like this that EUS-FNA has been optimized and gained widespread use for the benefit of patients. The indications for interventional EUS are currently expanding rapidly, with several diagnostic and therapeutic procedures emerging. EUS-guided transluminal internal biliary drainage, through the needle confocal laser endomicroscopy of pancreatic cysts, and recently, also EUS-guided gastroentero anastomosis, are examples of such evolving techniques [10]. The evidence of the majority of these procedures consists mostly of retrospective noncomparative studies, with a strong likelihood of selection bias. In order to optimize the results of these developing techniques for patients, we want to emphasize the urgent need for solid well-powered RCTs, not only comparing the interventional EUS procedures with current standard of care, but also to evaluate specific techniques and accessories.

Whereas EUS-FNA cytology of solid pancreatic lesions is fully sufficient when a diagnosis of malignancy is needed, several complex clinical conditions demand a more detailed analysis, which only histology can give. This might be the case with neuroendocrine lesions, autoimmune pancreatitis, lymphoma, subclassification of tumors, etc. In these cases, core biopsies are needed in order to achieve histological evaluation for subtyping. In addition, immunohistochemistry and genotyping may be required in the near future, both for individualized neoadjuvant treatment of pancreatic cancer and for tailored therapy in patients with inoperable or disseminated malignant disorders.

Several papers have evaluated novel needles and techniques with the aim of securing procurement of core specimens [11] [12]. At present, it is clear that a reverse beveled needle (ProCore; Wilson-Cook Medical Inc., Winston Salem, North Carolina, USA), promoted as a core needle, achieves neither a higher diagnostic sensitivity nor does it obtain more core specimens than a standard needle [6] [13]. Recently, a few other needles have been tested, demonstrating promising preliminary results. However, the few studies available at present are mostly of a retrospective nature and of low quality with a high risk of bias, and we believe that it is too early to finally conclude whether core tissue is better taken with these needles than with a standard needle. However, of note is a recent RCT by Bang et al., which assessed the 22-gauge Franseen fine-needle biopsy (FNB) needle (Acquire; Boston Scientific Corp.) vs. a standard 22-gauge needle for procurement of core biopsies from solid pancreatic lesions. The authors found that the median area of total tissue was 6.1 vs. 0.28 mm2, respectively (P < 0.001) [14]. Notably, immunohistochemistry was possible in 100 % vs. 68.3 %, respectively (P  < 0.001). In the study by Saxena et al., using standard needles, a core was obtained in 60.6 % in the slow-pull group and 46.7 % in the suction group [9], but other studies have shown higher procurement of tissue cores with standard needles [15]. Predictably, the wish for tissue acquired by EUS-FNA and EUS-FNB will steeply increase; hence, this is an area that needs further research, and currently special indications already warrant the need for new needle designs and techniques for sufficient tissue acquisition.

Further optimization of technical procedural issues with EUS-FNA of solid pancreatic lesions seems to have reached a plateau. Only further well-conducted studies will show whether new needle designs may optimize the diagnostic yield further. Only the future will show whether a 100 % sensitivity may be reached; however, the pancreas may not let us.