Diagnostic Accuracy of Preoperative Diagnostic Procedure in the Assessment of Lymph Node Metastases by NF-PanNENs

Overview

Pancreatic Neuroendocrine Neoplasms (PanNEN) are a heterogeneous group of neoplasms that arise from the endocrine cells of the pancreatic gland. Non-functioning (NF-PanNEN) represents the most common forms and do not produce syndromes due to hormonal hypersecretion. Several prognostic factors have been demonstrated for NF-PanNEN. The presence of nodal metastasis and lymph node ratio are widely considered predictors of disease-free survival and even the number of positive nodes has been found to be associated with recurrence. In addition to traditional imaging exams, diagnostic work-up should include a Positron Emission Tomography/CT with 68Ga labeled somatostatin analogues, which have shown a high sensitivity and specificity while 18F-FDG PET can be associated for evaluating the possible presence of a high-grade component. Moreover, pancreatic endoscopic ultrasound (EUS) is usually part of the preoperative staging both for imaging details and cytological sampling. Therefore an accurate identification of nodal metastases preoperatively may have important implications for the extent of surgical resection and lymphadenectomy and even for a prognostic outcome. In this study the investigators will evaluate prospectively the accuracy of these diagnostic exams in detecting the lymph node status. Patients with sporadic NF-PanNEN who are candidates for surgical resection will undergo CE-CT scan, 68Ga DOTATOC (and eventually 18F-FDG) PET/MRI and EUS with FNA/B. Reference standard for defining the presence of nodal metastases is represented by pathological examination on the specimen.

The study is designed as a prospective monocentric observational cohort study. The primary endpoint of the study is to assess the accuracy of preoperative diagnostic procedures in terms of sensitivity and specificity in detecting nodal metastases in sporadic NF-PanNEN as determined by pathological examination on the specimen. Patients with a suspect of a sporadic NF-PanNEN will be considered for the study at the time of the first visit. All consecutive patients who will meet inclusion criteria, after informed consent, will be included. The preoperative work-up will be performed in our center. All the examinations required for this study are necessary for the correct preoperative staging, representing the standard diagnostic work-up in patients with PanNEN. Computed tomography (CT) images will be obtained by multiphase technique. Non-enhanced, arterial and portal phase images will be obtained in all the patients. Computed tomography (CT) images will be obtained by multiphase technique. Non-enhanced, late arterial and portal phase images will be obtained in all the patients as well. Every lymph nodal finding with short-axis diameter \>1cm or with pathological contrast enhancement will be considered as pathological. Pancreatic endoscopic ultrasound (EUS) is a relatively recent endoscopic technique which combines the high-frequency ultrasound technology with the direct endoscopic vision, permitting a sensitive evaluation of pancreatic lesions (from the stomach or the duodenum), an accurate vascular staging and the assessment of pathologic lymph nodes related to the tumor. Nodes larger than 1 cm, with well demarked and/or rounded margins and with typical echoic/elastographic features will be reported as metastatic. Moreover, the additive value of EUS is the possibility to sample pancreatic masses and lymph nodes with a fine needle aspiration (FNA), to obtain a cytological diagnosis. A PET/MRI with 68Ga-DOTATOC is also necessary for the staging of the disease. A 18F-FDG PET will be performed according to the clinician's discretion, especially in the presence of large lesions and/or liver metastases. As soon as the preoperative assessment will be completed, a surgical resection will be proposed to the patient according to the international guidelines on the management of NF-PanNEN. A standard pancreatic resection will be performed according to the site and the extension of the lesion. Nodes will be classified according to the Japanese classification of the biliary tract cancers. The extension of lymphadenectomy will be related to the type of pancreatic resection. LN that will be resected apart from the pancreatic specimen, will be sent for pathological examination in a separated cup labelled by the surgeon. In case of preoperative detection of nodal involvement in a station not included in the standard lymphadenectomy, the nodes in this station will be resected to confirm or not the radiological finding. Reference standard for defining the presence of nodal metastases will be represented by the pathological examination. All data regarding the clinical aspect, the radiological reports and the pathological exam will be fully available only to the surgeon. Radiologist, nuclear medicine physician, and pathologist will be blinded to the radiological findings. Several retrospective studies 11-12 reported a prevalence of nodal metastases in PanNEN of about 40%, in case of tumors \> 2 cm. No data have been found on the sensitivity and specificity of the various examinations in detecting nodal involvement. Given this prevalence and assuming a good accuracy based on this institute experience, the simple size required to assess the sensitivity and specificity of the methods was found to be of 150 patients, considering a 95% confidence interval with a margin of error of 10%. A descriptive analysis will be performed of the data obtained. The aim will be to evaluate the sensitivity, the specificity and the positive and negative predictive values of the preoperative staging (CE-CT scan, EUS, 68Ga DOTATOC/18F-FDG PET/MRI). The comparison between the exams in terms of prediction of the nodal status will be calculated with the McNemar test. A logistic regression model will be used to assess the risk of nodal positivity at the pathological report when the preoperative imaging demonstrates the features of nodal metastases. OR and 95% confidence interval will be calculated. The statistical correlation between the risk of nodal positivity at the pathological report and some preoperative variables will be evaluated with the chi-square test and ROC curves analysis.