Feasibility of a 20-gauge ProCore needle in EUS-guided subepithelial tumor sampling: a prospective multicenter study

Gastrointestinal subepithelial tumors (SETs) include benign, potentially malignant, and malignant lesions. Endoscopic ultrasonography (EUS) is a useful tool for the characterization of SETs, providing details on the gastrointestinal wall structures in addition to morphologic features [1]. However, the differential diagnosis of hypoechoic SETs remains challenging, and a definitive diagnosis can rarely be established on imaging modalities alone [2]. As the treatment decision for patients with SETs largely depends on the histopathological diagnosis, effective and high-quality tissue acquisition is required.

In this context, EUS-guided sampling methods, including fine-needle aspiration (FNA), Trucut biopsy (EUS-TCB), and fine-needle biopsy (EUS-FNB) have been introduced [3]. EUS-FNA with a 22-gauge or 25-gauge needle provides cytological aspirates, with a diagnostic yield of 70% to 79% [4, 5]. Alternatively, tissue cores can be procured with EUS-TCB, which enables immunohistochemical analysis and thus facilitates a definitive diagnosis [6]. However, the diagnostic yield of EUS-TCB is not higher than that of EUS-FNA, and technical failures frequently occur [5, 7]. The Trucut needle is relatively stiff, and the firing mechanism produced by the torqued echoendoscope limits its maneuverability and accessibility in certain locations within the gastrointestinal tract, such as the gastric antrum or duodenum [8].

Recently, a novel EchoTip ProCore high definition ultrasound biopsy needle with a coiled sheath has become available. This needle, which has a unique reverse bevel technology, allows for a simultaneous core biopsy to be obtained along with aspirated material. The 19-gauge ProCore needle was shown to obtain histologically adequate samples in nearly 90% of cases, with excellent technical feasibility in both intra-intestinal and extra-intestinal mass lesions [9]. However, as this 19-gauge needle is still stiff, its maneuverability is not satisfactory in the gastric antrum or duodenum. Furthermore, the diagnostic yield of the 22-gauge ProCore needle was found to be 81.8% to 86.0% in gastric SETs [10–12]. The novel 20-gauge needle offers improved flexibility for those more difficult EUS-FNA biopsy approaches, with easy to-and-fro passage of the needle, along with the benefits of the larger 20-gauge needle to yield histologic grade tissue.

Although there is evidence supporting the utility of these procedures, diagnostic yields have varied according to the location in the gastrointestinal tract and specific needle types. The aim of this prospective multicenter study was to investigate the feasibility and diagnostic yield of EUS-FNB with a 20-gauge ProCore biopsy needle in the diagnosis of gastrointestinal SETs.

Of the 39 enrolled patients, three were excluded from the final analyses: two had lesions that were suspected to be larger than 2 cm, but were found to be < 2 cm on EUS examination, and one lesion was revealed to originate from the muscularis mucosa layer. Ultimately, 36 patients were included in the analysis (Fig. 2). Demographic and clinical characteristics of the SETs are shown in Table 1.

Age (years)	62.5 (27–82)
Male: female	19:17
Tumor size on EUS (cm)	2.5 (2.0–15.0)
Location
Esophagus	4 (11.1)
Stomach	30 (83.3)
Cardia	3 (10.0)
Fundus	4 (13.3)
Body	18 (60.0)
Antrum	5 (16.7)
Duodenum	2 (5.6)
Originating layer
Submucosa	6 (16.7)
Muscularis propria	27 (75.0)
Both layers	3 (8.3)
Final diagnosis
Gastrointestinal stromal tumor	20 (55.6)
Leiomyoma	7 (19.4)
Heterotopic pancreas	2 (5.6)
Schwannoma	1 (2.8)
Glomus tumor	1 (2.8)
Carcinoma	2 (5.6)
No final diagnosis	3 (8.3)

Variables are presented as number (%) or median (range)

EUS, endoscopic ultrasonography

Procedural characteristics and outcomes are summarized in Table 2. The needle punctures were successful in all cases irrespective of tumor location. The EUS-FNB results were diagnostic accuracy in 88.9% (32/36) of SETs. Tissue adequacy was optimal in 97.2% (35/36) of FNB specimens according to the endosonographers’ on-site visual evaluations, and in 88.9% (32/36) of specimens according to the pathologists. The mean numbers of needle passes required to obtain optimal tissue core samples were 1.2 ± 0.6 macroscopically and 1.5 ± 1.0 histologically. The median number of needle passes was one for both evaluations. The tissue adequacy was macroscopically optimal in 86.1% of cases after the first needle pass, and each needle pass yielded adequacy rates over 85% (Table 3). The tissue adequacy increased as the number of needle passes increased, with macroscopically optimal core samples being obtained in 94.4% of patients after two needle passes (Fig. 3). Three cases were considered as optimal by the endosonographer but were revealed to be suboptimal on histopathological examination: there was insufficient submucosal tissue in two cases and only fibrotic tissue was found in one case.

Diagnosis achieved	32 (88.9)
Sample adequacy during three needle passes
Macroscopic adequacy	35 (97.2)
Definite tissue core	32 (88.9)
Suspicious tissue core	3 (8.3)
Histological adequacy	32 (88.9)
Number of required needle passes to obtain optimal tissue core samplea
Macroscopic	1.2 ± 0.6
Histological	1.5 ± 1.0
Technical failure	2 (5.6)
Adverse events	2 (5.6)

Variables are presented as number (%) or mean ± standard deviation

^aIf the optimal core sample was not obtained after three needle passes, the number of passes required was considered to be four

	Pass 1 (n = 36)	Pass 2 (n = 36)	Pass 3 (n = 34)	Total (n = 36)
Macroscopic adequacy	31 (86.1%)	33 (91.7%)	32 (94.1%)	35 (97.2%)
Histological adequacy	27 (75.0%)	29 (80.6%)	27 (79.4%)	32 (88.9%)

Technical failure was encountered in two cases (5.6%) after two needle passes in the fundus and duodenal bulb. However, optimal tissue samples were obtained before the technical failure in both cases. In terms of adverse events, procedure-related bleeding occurred in two cases (5.6%). The bleeding was associated with needle puncture and was managed endoscopically with hemoclips.

Finally, 10 patients with diagnostic results avoided surgery: seven patients with leiomyoma, two with heterotopic pancreas, and one with glomus tumor. The diagnostic yield of EUS-FNB was 95.0% (19/20) in patients with a GIST. One lesion, which was not diagnosed on EUS-FNB, was surgically resected and confirmed as a GIST. Of 19 patients with a GIST, 13 had undergone surgical resection by the time of the analyses, and the histopathological diagnoses were consistent with the EUS-FNB diagnoses.

We investigated the diagnostic yield of EUS-FNB using a novel 20-gauge ProCore needle with a coiled sheath in the diagnosis of gastrointestinal SETs. Needle punctures were successful in all cases, irrespective of the location, and the diagnostic yield was 88.9%. The rates for obtaining macroscopically and histologically optimal core samples with EUS-FNB were 97.2% and 88.9%, respectively. The median number of needle passes for both macroscopic and histological adequacy was one. Our findings suggest that the 20-gauge ProCore needle may yield optimal core tissue samples with fewer needle passes compared with 22-gauge ProCore needle.

EUS-guided samplings are pivotal methods for tissue acquisition in gastrointestinal SETs, and a number of studies have evaluated the feasibility of EUS-guided sampling techniques in the diagnosis of SETs. EUS-FNA usually yields small sample volumes that are mainly processed for cytological evaluation [4, 5]. However, the cytological aspirate obtained by EUS-FNA is quantitatively low, and is often insufficient for differential diagnosis, especially in cases of gastric mesenchymal tumors that mandate an immunohistochemical assay. Samples with preserved tissue architecture are necessary to make a definitive diagnosis of hypoechoic SETs, especially when they are located in the muscularis propria layer. Although EUS-TCB provides large core tissue samples allowing histological examination as well as immunohistochemical staining, the Trucut needle is associated with technical difficulties because of its inherent stiffness, which results in a high technical failure rate [5, 6, 13]. Furthermore, the Trucut needle allows only one pass in a single axis, which thereby results in a limited diagnostic yield.

Recently, EUS-FNB technique has been developed, allowing core biopsy samples to be attained along with aspirated material. The ProCore needle is made of stainless steel, with a nitinol stylet and there is a reverse bevel to hook and cut tissue. Studies suggested that EUS-FNB may be advantageous for optimizing specimen adequacy and diagnostic accuracy. The 19-gauge ProCore needle showed histologic adequacy of 89.5% and diagnostic accuracy of 86.0% in the diagnosis of intra-intestinal and extra-intestinal mass lesions [9]. The diagnostic accuracy was 81.8–86.0% for gastric SETs when the 22-gauge needle was used [10–12]. In the present study, optimal macroscopic and histological core samples were procured in 97.2% and 88.9% of cases with three needle passes, which resulted in a high diagnostic histologic accuracy rate. Furthermore, adequate tissue core was obtained within two needle passes in most cases, with only 5.6% of cases requiring three needle passes to achieve a diagnosis.

The diagnostic yield of EUS-guided sampling depends on a variety of factors, such as the nature of the target lesion, site of the puncture, the availability of a cytopathologist, the experience of the endosonographer, and the type and size of the needle used [3]. Regarding the needle size, the large-caliber needles seem to have the advantage of acquiring more tissue, which enables the assessment of architectural features. However, a larger needle is prone to have technical difficulties with respect to its maneuverability and accessibility, whereas a smaller needle is flexible and can be fanned in multiple directions within the target lesion. Indeed, studies comparing the diagnostic performance for SETs did not demonstrate any significant advantage of EUS-TCB or EUS-FNB over a standard FNA needle in terms of cytologic parameters, amount of diagnostic cell block material, adequacy, and accuracy [4, 5, 14]. Of note, the number of needle passes required for diagnosis was significantly lower when using the ProCore needle, suggesting that a better quality sample was obtained in each pass [11, 14–16]. In the present study, the median number of needle passes to achieve both macroscopic and histological adequacy was one. The first pass of the 20-gauge ProCore needle yielded a histologically optimal tissue core in 75.0% of cases, and the histological tissue adequacy on each pass is over 75% throughout the procedures. These results demonstrate the high quality of the tissue obtained by a single pass of the ProCore needle.

It is sometimes difficult to determine whether sufficient core tissues are obtained during EUS-guided sampling, as indicated by the discrepancy between macroscopically and histologically assessed tissue adequacy. One possibility is that the visible materials do not consist of a tissue core, while another is that the materials acquired are not representative of the target lesion. The latter can be more challenging in clinical practice, because a tissue core may well be acquired but then revealed to give little information in respect to the diagnosis. In this present study, three cases were considered suboptimal for histological evaluation despite being macroscopically assessed as optimal core samples. All three cases contained visible tissue materials that seemed to be core samples, but were later revealed to be non-diagnostic. As most institutions do not have on-site pathologists, certain criteria for the macroscopic visual assessment of a specimen by the endosonographer can be helpful to ensure the adequacy of tissue cores and to reduce unnecessary punctures.

Although bleeding and perforation are potentially life-threatening adverse events of EUS and EUS-guided procedures, the incidence of adverse events has been reported as being low [17, 18]. In addition, most adverse events were caused by 19-gauge needles [7, 19]. Previous report using 22-gauge ProCore needles showed either no or low adverse event rates [10–12, 20]. In the present study, minor bleeding occurred during the procedure in two cases (5.6%), and was controlled endoscopically, thereby supporting the safety of EUS-FNB procedures.

Our study had several limitations of note. First, we included only academic centers with highly experienced endosonographers. Second, the size of all SETs included in our study was ≥2 cm. According to recent guidelines for SETs, when neoplastic SETs are 2–5 cm in diameter or when SET < 2 cm have clinically malignant features on endoscopy, the guidelines recommend detailed examination with EUS, computed tomography with contrast enhancement, and/or EUS-FNA. Clinically malignant features means irregular borders, ulceration, and/or growth during endoscopic follow-up. When there are no clinically malignant features, gastric SETs < 2 cm could be followed up by endoscopy or EUS once or twice a year until the tumors increase in size or become symptomatic, even if they are diagnosed as GISTs later on [21]. Therefore, we included only SETs with ≥2 cm in size. Third, in cases which do not need treatment surgically or endoscopically, the FNB results were determined as final diagnosis. There is a very rare possibility of malignant transformation in benign tumors, but it is a problem related to the natural history of tumors, not a wrong diagnosis. Fourth, cytological aspirates and histological core samples were not interpreted separately. In previous studies, the amount of diagnostic cell block material did not vary according to the use of either a beveled or standard needle, and the use of a beveled needle provided no benefit in terms of the diagnostic cell block [14]. Another possible limitation of our study is that the mitotic count and Ki-67 labeling index of the GIST were not determined. Although the diagnosis of GIST was successfully made before surgical resection, there may be considerable discrepancy in the mitotic count or Ki-67 index of the tumors between the EUS-FNB and surgical specimens [12].

In conclusion, EUS-FNB using a 20G ProCore needle is a technically feasible, safe, and effective modality for histopathologic diagnosis of gastrointestinal SETs larger than 2 cm, providing adequate tissues core with fewer needle passes.