Ethics approval and patient consent
The study was approved by the Wakayama Medical University Ethics Committee (IRB No: 2135). Informed consent will be obtained from all patients. The trial was registered with the University Hospital Medical Information Network (trial registration No.: UMIN000030634). This protocol was prepared in conformance with the Standard Protocol Items: Recommendations for Interventional Trials (SPIRIT) guidelines [see Additional file 1].
Study aims and design
The major objective of this study is to determine whether the Franseen needle is superior to the conventional needle for accurate diagnosis of malignancy in patients with pancreatic mass. This multicenter randomized controlled study will compare 22G conventional and Franseen needles in terms of the efficacy and safety of EUS-FNA for evaluating the malignancy of mass lesions in the pancreas. As this comparative study mainly aims to assess the specimens obtained with the first puncture, no special protocol will be established regarding the technique for the second or subsequent punctures. The flowchart in Fig. 1 illustrates the recruitment, allocation, and follow-up process. The primary endpoint in this study will be diagnostic accuracy for malignancy, whereas secondary endpoints will include (i) sensitivity and specificity for diagnosis of malignancy, (ii) procedure completion rate, (iii) sample cellularity, and (iv) incidence of complications.
A total of 520 patients will be recruited from 21 endoscopic centers in Japan.
Registration will be performed at each participating center following local ethics committee approval of the study protocol and of the informed consent documentation and forms.
The following inclusion criteria will be applied: (i) age ≥ 20 years; (ii) pancreatic mass lesion detected on diagnostic imaging; (iii) performance status ≤2; (iv) indication for histological evaluation to predict the clinical course and to select treatment methods; (v) written consent obtained following adequate explanation of the study aims, design, and procedures.
The following exclusion criteria will be applied: (i) blood vessel or other tumors located between the lumen of the gastrointestinal tract and the target lesion; (ii) bleeding tendency, defined as an international normalized ratio of the prothrombin time >1.5 or platelet counts <50000 cells/µl; (iii) any condition expected to hinder endoscope insertion; (iv) gastrointestinal reconstruction after gastrectomy; (v) serious complication involving another organ; (vi) known diffuse autoimmune pancreatitis; (vii) any other condition or situation determined by a study investigator to represent reason for ineligibility.
Registration of candidates
At each participating center, the on-site study investigators will obtain informed consent from the candidates and use an electronic data capture system to confirm that the candidates meet the eligibility criteria (i.e., the candidates meet all the inclusion criteria and none of the exclusion criteria), input the necessary information, and register the candidates with the registration secretariat. After confirming that the candidate meets the criteria, a registration number will be issued, and the registration will be considered complete. If any required input data are missing or the candidate does not meet the eligibility criteria, no registration number will be issued, and registration will be deemed incomplete. For each candidate, the date of registration is defined as the date when the registration number was issued by the registration secretariat via the electronic data capture system.
Randomization and blinding
Each subject will be randomly assigned to either of the treatment methods at a ratio of 1:1 according to a web-based registration program system (based on dynamic allocation by the minimization method). At the time, the following factors will be the adjustment factors for assignment to prevent large bias (diameter of tumor, site where the tumor is located, pattern of contrast CT, and institution.) The registration secretariat will strictly control the program to prevent leaking of the assignment information to outside. Blinding will not be used in this study.
The procedures will be conducted by experienced operators who have performed at least 100 EUS-FNA procedures prior to the initiation of the study.
An EUS device with a convex transducer will be used to observe the target lesions through the lumen of the gastrointestinal tract. In Doppler mode, the EUS device will be used to check whether any large blood vessel crosses the planned puncture route.
The maximum diameter of the lesion in the direction of the puncture route will be measured in order to determine the position of the puncture needle stopper. Puncture will be performed after retracting the stylet by approximately 5 mm.
While applying negative pressure using a 10-mL syringe and observing the puncture needle under ultrasound guidance in real time, the needle will be moved back and forth within the lesion approximately 20 times.
Processing of specimens
The stylet will be re-inserted into the needle to eject the collected tissues from inside the needle. The specimens will be submitted in a container of formalin.
It is not necessary to use the same needle for subsequent punctures. The operator may freely select the puncture needles for subsequent punctures depending on the size and number of specimens obtained during the first puncture. In this study, only the specimen obtained with the first puncture will be used to evaluate the pre-specified outcomes. The EUS-FNA procedure will be considered complete when the operator determines that sufficient specimens have been collected at the second or subsequent punctures
Histopathological diagnosis will not be based on cytology but on histology. Specimens fixed with formalin will be embedded in paraffin blocks. Immunostaining and nuclear staining will be performed as needed. In this study, the specimens obtained at the first puncture will be sent to Hoken Kagaku West Japan Co., Ltd., who will prepare the specimens and send them to Kyoto Prefectural University of Medicine for diagnosis. Two experienced pathologists who have performed at least 1000 cytology and histology evaluations of EUS-FNA specimens will assess the specimens independently. The two pathologists will be blinded to the type of needle used. Histology evaluation will be performed to determine: (i) sample cellularity, (ii) preservation of tissue architecture, and (iii) histologic diagnosis. Sample cellularity is classified as rich (≥5000 cells), moderate (100–5000 cells), or poor (<100 cells).
Data will be collected using a standardized data entry form and entered into the data management system. Figure 2 provides an overview of the types of data collected and the timing of data collection.
For the purpose of this study, diagnosis is defined as malignant or benign mass. For patients receiving surgical resection, the final diagnosis will correspond to the pathological diagnosis established based on the resected sample. For patients with non-resection, the final diagnosis will be made based on the results of repeat EUS-FNA biopsy, diagnostic imaging such as computed tomography, magnetic resonance imaging, and positron-emission tomography, and observation of the clinical course for at least one year.
For the purpose of the present study, all patients will be followed-up for at least one year after EUS-FNA.
Sample size calculation
There have been no reports on the rate of correct diagnosis of pancreatic masses evaluated by EUS-FNA using the 22G Franseen needle. However, a previous study focused on puncture needles designed to sample sufficient tissue for core biopsy reported that the rate of correct diagnosis was 75% for the conventional needle (EchoTip Ultra needle), compared to 80% for the specially designed EchoTip ProCore needle . Assuming that the rate of correct diagnosis is higher when using the Franseen needle than when using the EchoTip ProCore needle, we expect the rate of correct diagnosis of pancreatic mass malignancy to be 85% when EUS-FNA is performed using the Franseen needle. A sample size of 500 participants (250 per group) would provide a power of 80% (1-β) considering a significance level of α = 0.05 (chi-square test) for between-group comparisons of diagnostic rate. Assuming that approximately 10 participants per group will have to be considered inadequate for inclusion in the final analysis, the target sample size in this study is established at 520 participants (260 per group).
Major analysis and judgment criteria
The primary objective of this study is to determine whether the test diagnostic device (22G Acquire™ needle) is superior to the conventional diagnostic device (22G Expect™ needle) for accurate diagnosis of malignancy. If the accuracy of diagnosis is significantly higher in the Acquire™ 22G needle group than in the 22G Expect™ needle group, it will be concluded that the test diagnostic device (22G Acquire™ needle) is a more promising diagnostic device. If the difference is not significant, it will be concluded that the 22G Acquire™ needle is not a superior diagnostic device compared with the 22G Expect™ needle. The main analysis will use the Mantel-Haenszel test because adjustment factors other than the site are used for stratification in the full analysis set (FAS). When comparing the two groups in terms of the rate of correct diagnosis, the difference will be considered significant if the p-value in the Mantel-Haenszel test is below the significance level taken into consideration in the power analysis (a = 0.05). The Clopper-Pearson exact test will be used to estimate the 95% confidence interval of accuracy for the diagnosis of malignancy in each group. In addition, logistic regression analysis with adjustment for covariates (other than site) will be used to calculate the odds ratio (with 95% Wald confidence interval) for the outcome associated with using the 22G Acquire™ needle versus using the 22G Expect™ needle.
Secondary efficacy analysis
Sensitivity and specificity for the diagnosis of malignancy
Sensitivity and specificity (with the Clopper-Pearson exact 95% confidence interval) will be estimated for each group in the FAS. The chi-square test will also be conducted, and the odds ratio (with 95% confidence interval) will be calculated to evaluate the between-group differences in sensitivity and specificity. The adjusted odds ratio (with 95% confidence interval) will be calculated using logistic regression analysis including the assignment factors.
Procedure completion rate
Procedure completion rate (with the Clopper-Pearson exact 95% confidence interval) will be estimated for each group in the FAS. The chi-square test will also be conducted and the odds ratio (with 95% confidence interval) will be calculated to evaluate the between-group differences in procedure completion rates. The adjusted odds ratio (with 95% confidence interval) will be calculated using logistic regression analysis including the assignment factors.
Sample cellularity will be tabulated by grade for each group in the FAS. The chi-square test will also be conducted to evaluate the between-group difference in tissue sampling rate. The adjusted odds ratio (with 95% confidence interval) will also be calculated using a proportional odds model including the assignment factors.
Incidence of complications
The incidence of complications (with the Clopper-Pearson exact 95% confidence interval) will be estimated for each group in the FAS. The chi-square test will also be conducted and the odds ratio (with 95% confidence interval) will be calculated to evaluate the between-group differences in the incidence of complications. The adjusted odds ratio (with 95% confidence interval) will be calculated using logistic regression analysis including the assignment factors.
The severity of adverse events will be graded according to the guidelines issued by the American Society for Gastrointestinal Endoscopy . The incidence rate and proportion of adverse events during follow-up will be calculated for the safety analysis set, both overall and stratified by severity. Clopper-Pearson exact 95% confidence intervals will be estimated for these occurrence proportions.
Exploratory subgroup analyses will be conducted to investigate the interaction between treatment effects and background factors. In principle, median values will be used for the stratification of the study sample into subgroups according to each factor of interest. Since the subgroup analyses are not statistically powered and no multiplicity adjustment is performed, the results of these analyses will only be interpreted as exploratory. In subgroup analyses, the odds ratio will be used as a summary of therapeutic effects. A Forest plot will be used to summarize the point estimate and 95% confidence interval for each subgroup. The factors and cutoff values to be used for exploratory analysis will be established in a separate statistical analysis plan.