A total of 89 UR-PC patients who underwent EUS-FNA or EUS-FNB (EUS-FNAB) using 22-G needles between December 2018 and December 2021 at three hospitals in Japan (Fukushima Medical University Hospital, Aizu Medical Center of Fukushima Medical University, and Ohtanishinouchi Hospital) were enrolled (Fig. 1). Among these patients with UR-PC, 56 were followed up or evaluated for MSI. A total of 23 and 33 patients underwent EUS-FNB and EUS-FNA, respectively. In contrast, 33 patients were excluded because they were not evaluated for MSI: 14 patients were treated with best supportive care, 13 were treated with chemotherapy at other hospitals, and six chose not to undergo MSI evaluation. Thirteen patients who received treatment at other hospitals did not undergo subsequent MSI evaluation of the EUS-FNAB specimens.

All institutions that participated in this study were affiliated hospitals at the Fukushima Medical University. Therefore, the procedural and histological diagnostic criteria were almost identical. In addition, histological specimens obtained from the Ohtanishinouchi Hospital were sent for analysis to the Department of Diagnostic Pathology, Fukushima Medical University, through highly secure internet communication.

After the patients were sufficiently sedated with intravenous administration of midazolam and placed in the left lateral decubitus position, we gently inserted an echoendoscope. When the lesion was observed on the monitor, the lack of blood flow on the puncture line was confirmed using doppler mode. The lesion was punctured and the stylet was removed. The needle was passed back and forth 20 times in the lesion while suction was applied using a 10-mL or 20-mL syringe. White lumbricoid specimens were separated from all the specimens collected using EUS-FNAB. The lumbricoid specimens were preserved in a bottle filled with formalin for subsequent histological diagnosis. The remaining specimens were used for cytology and rapid onsite evaluation (ROSE).13 Repeat aspiration was performed until a sufficient specimen was obtained, as confirmed by ROSE. We usually complete EUS-FNAB after four needle passes in accordance with the report by Suzuki et al.14 However, when sufficient specimens were not confirmed by ROSE, we performed additional needle passes.

The following echoendoscopes and ultrasonography equipment were used: GF-UC240AL-5, GF-UCT260, EU-ME-1, and EU-ME-2 (Olympus Medical Systems, Tokyo, Japan). The needles were selected as follows. For patients with vessels near the puncture line, a conventional FNA needle was used because of its good penetration ability. Otherwise, the endoscopist randomly selected the needle. The Franseen Acquire 22-G needle (Boston Scientific, Marlborough, MA, USA) was used for EUS-FNB (Fig. 2). The cutting area of the Franseen needle was larger than that of a conventional needle. The needles used for EUS-FNA were an EchoTip 22-G (Cook Medical Inc., Winston Salem, NC, USA), EZ Shot 3 Plus 22-G (Olympus Medical Systems), and Expect 22-G (Boston Scientific).

All procedures were performed by pancreaticobiliary specialists who had performed more than 200 EUS-FNAB procedures or by beginners under the guidance of pancreaticobiliary specialists.

Patient characteristics (age, sex), UR-PC data (tumor size, location, and progression), and procedural outcomes (puncture route, puncture number, possibility of histological diagnosis, possibility of performing MSI evaluation, and adverse events) were compared between patients who underwent EUS-FNB and those who underwent EUS-FNA. Additionally, a multivariate analysis was performed to determine the factors that influenced the possibility of MSI evaluation. Tumor size was measured using computed tomography or EUS. Locally advanced UR-PC was determined under the following conditions: (1) invades/contacts the portal vein or superior mesenteric vein by >180° with a range of invasion/contact transcending the inferior duodenal angle; (2) invades/contacts the celiac artery or superior mesenteric artery by >180°; (3) invades/contacts the common hepatic artery with a range of invasion/contact extending to the celiac artery or proper hepatic artery; and (4) invades/contacts the aorta.15,16 In contrast, UR-PC was diagnosed by histology or cytology using EUS-FNAB. Cytology was classified on a scale of classes I to V (I, normal; II, atypical but benign; III, difficult to diagnose as benign or malignant; IV, suspected malignant; and V, malignant). Malignancy was defined as a class IV or V cytology.

MSI evaluations for the three hospitals were performed by an external agency (Falco Holdings Co., Kyoto, Japan). If a histological specimen did not satisfy the conditions required for MSI evaluation, the specimen was not submitted to the outside agency for pathological judgment.

Histological diagnosis was made using all specimens submitted to the department of pathology. A formalin-fixed paraffin-embedded (FFPE) block was designated for each procedure performed at the Fukushima Medical University and Aizu Medical Center. Alternatively, an FFPE block was prepared for each puncture at the Ohtanishinouchi Hospital. To diagnose whether a tumor is malignant, the specimen must include tumor tissue that has not degenerated or crushed. Furthermore, when a specimen contains inflammatory contents, it becomes difficult to identify whether atypical cells are tumoral or reactive to inflammation.

The conditions for histological specimens required by the outside agency (Falco Holdings Co.) were as follows: more than 2,000 PC cells to extract DNA for MSI evaluation, PC cell content >50%, and more than five unstained PC slide specimens must be prepared. An MSI detection kit, called the Promega panel, was used to analyze MSI using five mononucleotide markers (BAT-25, BAT-26, NR-21, NR-24, and MONO-27). Although the Bethesda panel has been used for MSI evaluation for a long time,17 the Promega panel is superior to the Bethesda panel for MSI evaluation.18 In addition, unlike the Bethesda panel, the Promega panel does not require a normal tissue control.19-21 When more than two markers are MSI positive, MSI-high PC is diagnosed.

The Mann-Whitney U-test was used to compare continuous variables that did not conform to a normal distribution with ordinal variables. The Fisher exact test was used to compare nominal variables. Logistic regression was used to investigate the factors influencing the possibility of MSI evaluation. Statistical significance was set at p<0.05. EzR (Saitama Medical Center, Jichi Medical University, Saitama, Japan)22 was used for all statistical analyses.

This multicenter retrospective study was approved by the Institutional Review Board of Fukushima Medical University (IRB No: 2021-002). Patients were not required to provide informed consent because this study used anonymous clinical data obtained after each patient had agreed to medical activities by written consent. For full disclosure, the details of this study are published on the homepage of the Fukushima Medical University.

Patient characteristics, tumor size, PC location, and PC progression were not significantly different between patients who underwent FNB and those who underwent FNA (Table 1). Regarding the outcomes of EUS-FNAB, the puncture route, puncture number, and adverse events were not significantly different between the two groups (Table 2). Regarding FNA needles, Expect was used in 20 patients, EZ Shot 3 Plus was used in nine patients, and Echotip was used in four patients. Gastric bleeding at the puncture site was observed in one patient who underwent FNA. Bleeding was treated with endoscopic hemostasis using clips. The possibility of histological diagnosis was also not significantly different between the two groups (FNB 91.3% [21/23] vs. FNA 72.7% [24/33]; p=0.10). The possibility of MSI evaluation was significantly more frequent with FNB than with FNA (82.6% [19/23] vs. 45.5% [15/33], respectively; p<0.01). An MSI-high tumor was observed in only one patient who underwent FNA.

Three needles were used in the patients who underwent EUS-FNA. However, Expect was used in most patients who underwent EUS-FNA. When patient characteristics and outcomes were compared between EUS-FNB and FNA using Expect, the results were similar to those of the comparison between EUS-FNB and FNA in all patients (Table 3). The possibility of histological diagnosis was not significantly different between FNB and FNA using Expect (91.3% [21/23] vs. 70.0% [14/20], respectively; p=0.12). The possibility of MSI evaluation using Expect was significantly higher in the FNB group than in the FNA group (82.6% [19/23] vs. 40.0% [8/20], respectively; p<0.01).

In the univariate analysis, EUS-FNB was a significant factor influencing the possibility of MSI evaluation. The current guideline recommends three to four needle passes with FNA needles or two to three needle passes with FNB needles.23 Therefore, EUS-FNB and greater than three punctures were included in the multivariate analysis. As a result, EUS-FNB was the only significant factor influencing the possibility of MSI evaluation in the multivariate analyses (Table 4).