Background: Prognosis of pNENs depends on staging and grading, which is based on the cyto-histological Ki67 labelling. EUS-FNA is considered the gold standard tech nique to obtain a cytological specimen in pre-therapeutic setting that can be used to evaluate ki67. Aim: The aim of this study was to establish the diagnostic accuracy of preoperative EUS-FNA Ki67 evaluation in a cohort of pNEN patients. Methods: This is a retrospective study from a prospectively collected database on patients who underwent surgery for pNENs from 2006 to 2019 and EUS-FNA Ki67 labelling. EUS FNA and surgery Ki67 (eKi67 and sKi67, respectively) values and grading, according to 2017 WHO classification, were compared and the diagnostic accuracy of EUS-FNA was evaluated with sKi67 as gold standard. eKi67 and sKi67 correlation was evaluated by Pearson’s index. Results: 112 pNEN patients enrolled. Correlation between eKi67 and sKi67 values was good (coefficient rZ 0.78, Figure 1). On EUS-FNA specimens 56/112 (50%) patients were classified as G1, 53 (47.3%) as G2 and 3 (2.7%) as G3, while on surgery they were respectively 59 (52.7%), 50 (44.6%) and 3 (2.7%). In 9.8% of patients grade was increased from G1 to G2 by surgical histology, while in 12.5% it was diminished from G2 to G1. No misclassification occurred in G3 patients (Table 1). Considering only patients with small tumours (<2 cm), similar misclassification rate was observed respectively in 9.5% (4/48) and 21.4% (nZ9) of cases. Sensitivity, specificity, positive and negative predictive values and accuracy of eKi67 to correctly classify G2 patients were respectively 78.4%, 77.4%, 73.6%, 81.4% and 77.7%. No predictive factors of misclassification were found at multiple regression analysis. Conclusions: This study represents the largest cohort of surgical pNEN patients with preoperative eKi67 evaluation. We found a good correlation between eKi67 and sKi67, but about 20% of patients are not correctly allocated in grading classes. This should be carefully considered especially in small tumours undergoing observation.
DIAGNOSTIC ACCURACY OF ENDOSCOPIC ULTRASOUND-FINE NEEDLE ASPIRATION (EUS-FNA) IN THE EVALUATION OF PANCREATIC NEUROENDOCRINE NEOPLASMS (PNEN) GRADING / Tacelli, M; Petrone, Mc; Capurso, G; Rossi, G; Archibugi, L; Testoni, S; Muffatti, F; Andreasi, V; Partelli, S; Doglioni, C; Falconi, M; Arcidiacono, Pg. - In: GASTROINTESTINAL ENDOSCOPY. - ISSN 0016-5107. - 91:6S(2020), pp. AB199-AB199.
DIAGNOSTIC ACCURACY OF ENDOSCOPIC ULTRASOUND-FINE NEEDLE ASPIRATION (EUS-FNA) IN THE EVALUATION OF PANCREATIC NEUROENDOCRINE NEOPLASMS (PNEN) GRADING
Testoni S;
2020-01-01
Abstract
Background: Prognosis of pNENs depends on staging and grading, which is based on the cyto-histological Ki67 labelling. EUS-FNA is considered the gold standard tech nique to obtain a cytological specimen in pre-therapeutic setting that can be used to evaluate ki67. Aim: The aim of this study was to establish the diagnostic accuracy of preoperative EUS-FNA Ki67 evaluation in a cohort of pNEN patients. Methods: This is a retrospective study from a prospectively collected database on patients who underwent surgery for pNENs from 2006 to 2019 and EUS-FNA Ki67 labelling. EUS FNA and surgery Ki67 (eKi67 and sKi67, respectively) values and grading, according to 2017 WHO classification, were compared and the diagnostic accuracy of EUS-FNA was evaluated with sKi67 as gold standard. eKi67 and sKi67 correlation was evaluated by Pearson’s index. Results: 112 pNEN patients enrolled. Correlation between eKi67 and sKi67 values was good (coefficient rZ 0.78, Figure 1). On EUS-FNA specimens 56/112 (50%) patients were classified as G1, 53 (47.3%) as G2 and 3 (2.7%) as G3, while on surgery they were respectively 59 (52.7%), 50 (44.6%) and 3 (2.7%). In 9.8% of patients grade was increased from G1 to G2 by surgical histology, while in 12.5% it was diminished from G2 to G1. No misclassification occurred in G3 patients (Table 1). Considering only patients with small tumours (<2 cm), similar misclassification rate was observed respectively in 9.5% (4/48) and 21.4% (nZ9) of cases. Sensitivity, specificity, positive and negative predictive values and accuracy of eKi67 to correctly classify G2 patients were respectively 78.4%, 77.4%, 73.6%, 81.4% and 77.7%. No predictive factors of misclassification were found at multiple regression analysis. Conclusions: This study represents the largest cohort of surgical pNEN patients with preoperative eKi67 evaluation. We found a good correlation between eKi67 and sKi67, but about 20% of patients are not correctly allocated in grading classes. This should be carefully considered especially in small tumours undergoing observation.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
