What is the best method for endoscopic ultrasound-guided fine needle aspiration? Needle types and aspiration techniques

Abstract

Introduction

Endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) has become a standard procedure for sampling many types of lesions within the gastrointestinal tract and adjacent organs for cyto-histopathological diagnosis.1 A 25-gauge (25-G), 22-G, or 19-G needle is commonly used for EUS-FNA.2 Many factors can affect the diagnostic accuracy of EUS-FNA.3 One major factor is the needle itself. To obtain adequate specimens, many types of EUS-FNA needles have been developed.4,5 Needles are mainly divided into two types: fine needles and reverse-beveled. Fine needles consist of soft needles and stiff needles. In addition, the usefulness of the slow-pull technique has recently been reported.6,7 Studies comparing the clinical impact of different needle types, needle sizes, and aspiration methods are needed. However, because many biases exist in clinical studies of humans, particularly in terms of individual differences among participants, results are largely dependent on the characteristics of the patients and tumors.8 The aim of this study was thus to evaluate EUS-FNA needles and aspiration techniques using animal and artificial models.

Methods

Protocol 1

We performed EUS-FNA on a pig liver through the esophagus under general anesthesia using a stylet. All types and sizes of needles traversed the same area of the liver to and fro 10 times, covering a distance of 10 mm with negative pressure applied using a 20-mL syringe or the slow-pull technique. A total of six punctures were performed for each needle by each of the endosonographers. The slow-pull technique involves low negative pressure by pulling the stylet out slowly. To evaluate the slow-pull technique, we used all sizes of stiff fine needles and reverse-beveled needles. All the obtained specimens were fixed in formalin for the cell block method. The specimens were scored according to the our own grading system (Fig. 2).

Figure 2

Grading of the obtained specimens. The correlation of grading points and reference figures are shown.

Protocol 2

EUS-FNA was performed using three materials: Japanese sweet bean jelly, tofu, and cow liver (Fig. 3). Again, we used three types of needles: stiff fine needle, soft fine needle, and reverse-beveled needle. Needle sizes were 22 G and 19 G. Two of the three endoscopists performed EUS-FNA with negative pressure applied using a 20-mL syringe or the slow-pull technique. During each puncture, the needle traversed all three materials to and fro 10 times, covering a distance of 10 mm. A total of three punctures were performed for each needle or method by each of the two endosonographers. All the obtained specimens were placed on the dish one by one. The FNA specimens were evaluated macroscopically and compared with each other.

Figure 3

Photographs of Japanese sweet bean jelly, tofu, and cow liver.

Results

Protocol 1

Three skilled endoscopists performed EUS-FNA using all types and sizes of needles with 20 mL of negative pressure and the slow-pull technique. All procedures were performed successfully. Visualization of the tip of the needle was similar for all kinds of needles, and puncture difficulties were also similar in the pig model. Each type of 25-G needle after puncture is shown in Fig. 1. The soft-type needle easily became bent after only one puncture. The mean ± standard deviation scores for specimens obtained with each type of needle (all sizes of needle included) are shown in Table 1. As a result, the mean score for reverse-beveled needles (4.1 ± 1.41) was significantly higher than that for soft-type needles (3.5 ± 1.79; P < 0.05, Dunn’s test). The results for mean score by needle size (all types of needle included) are shown in Table 2. The score for each size of needle showed no significant difference, even between 25 G and 19 G. The mean scores for all needles are shown in Table 3. The 25-G reverse-beveled needle (4.5 ± 0.54) provided significantly better scores than the 22-G soft needle (2.2 ± 1.17; Dunn’s test, P < 0.05). The most favorable method for core specimens was a 19-G stiff needle with 20 mL of negative pressure; the core specimen rate was 50%. However, there was no significant difference statistically. Next, we examined the usefulness of the slow-pull technique on pig liver under general anesthesia. The scores for 20 mL of negative pressure and the slow-pull technique are shown in Table 4. No significant difference was seen between 20-mL negative pressure and the slow-pull technique using either a stiff needle or a reverse-beveled needle.

Figure 1

Photograph of the 25-gauge needles after puncture. Blue, soft-type needle; black, stiff-type needle; gray, reverse-beveled needle.

Protocol 2

The aim of the second examination was to evaluate the usefulness of the slow-pull technique under different situations. EUS-FNA was performed on the experimental table using 22-G and 19-G versions of all types of needles. The results are shown in Fig. 4. Typical findings of specimens obtained using 22-G needles are also shown in Fig. 4. Comparing the slow-pull technique with 20-mL negative pressure, cow liver provided a small specimen but less blood, tofu provided less specimen, and Japanese sweet bean jelly provided a similar specimen to both.

Figure 4

Typical findings with the slow-pull technique and 20 mL of negative pressure. From the left, specimens of Japanese sweet bean jelly, tofu, and cow liver are shown.

Discussion

This is the first report to describe the effects of needle size, needle type, and the technique for providing negative pressure using a live animal model and artificial models on the experimental table. Some doctors have reported the usefulness of 25-G needles for FNA and 19-G needles for FNA biopsy.5,9–20 The usefulness of the reverse-beveled needle and the slow-pull technique has recently been reported.4,5,21 The possibility of reduced bloody aspirates using a 25-G needle and the possibility of obtaining adequate amounts of specimen by a 19-G needle or a reverse-beveled needle may increase diagnostic accuracy.9,11,21

The slow-pull technique provides low-grade negative pressure and may produce less bloody specimens to increase diagnostic accuracy.6

However, diagnostic accuracy might be limited by several factors including individual differences, such as anatomical location, tumor size, tumor necrosis, tumor fibrosis, grade of tumor malignancy, vascularity, surrounding tissue, and so on. These limitations could lead to controversial results in clinical studies of humans. We therefore planned to evaluate the best type of needle, the best size of needle, and the best method of providing negative pressure under stable conditions. To eliminate individual differences, we prepared a live pig model under general anesthesia in Protocol 1, and models using Japanese sweet bean jelly, tofu, and cow liver in Protocol 2. We used liver to represent bloody tumor, tofu for soft tumor, and Japanese sweet bean jelly for hard tumor.

First, we will discuss the results for Protocol 1. Visualization of the needle tip and the difficulty of EUS-FNA in the live pig model showed no differences in needles under normal conditions as in this study. Needle size, needle type, and technique of negative pressure were not significantly associated with provision of better specimens (Tables 1 and 2). The needle obtaining the best score was the 25-G reverse-beveled needle with 20 mL of negative pressure (Table 3). The specimens obtained with the 25-G needle provided less bloody samples, so less blood may be one factor associated with better specimens in cell block analysis. However, 20 mL of negative pressure and the slow-pull technique showed no differences with the different types and sizes of needle in the liver. Rather, the slow-pull technique tended to be associated with lower scores than 20 mL of negative pressure (Table 4). Next, we confirmed the usefulness of the slow-pull technique under some conditions in Protocol 2. The slow-pull technique may avoid blood contamination in liver samples, but in less bloody tumors, 20 mL of negative pressure is better, as for the tofu in Fig. 4. Both techniques for providing negative pressure are useful for harder tumors, as in the Japanese sweet bean jelly (Fig. 4). For hard tumors like Japanese sweet bean jelly, negative pressure may not be needed.

Our study results were obtained under stable conditions with no effects of individual differences. Our results may thus represent objective findings without external effects. Conversely, our results may not sufficiently reflect clinical findings. These are limitations in our study.

In summary, the score for the reverse-beveled needle was better than the score for the soft-type needle. The size of the needle and the slow-pull technique did not affect FNA specimens in our model. The slow-pull technique may be associated with reduced blood contamination and may produce better specimens for cell block analysis. By contrast, small specimens may be obtained from less bloody tumors. Negative pressure may not be useful in a hard solid tumor. When we perform EUS-FNA, needles and conditions should be selected based on the relevant patient and tumor characteristics.

Article information

References