Évaluation du grade des tumeurs endocrines du pancréas et ponction sous échoendoscopie
G. Monges,F. Figuereido,M. Giovannini
DOI: https://doi.org/10.1007/s10190-010-0015-1
2010-02-01
Acta Endoscopica
Abstract:IntroductionPancreatic endocrine tumors (PETs) do not have the same prognosis and differ in their clinical behavior. If it is possible to diagnose PETs and to determine their malignant potential through specimens obtained by endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA), it will be of great importance in the management of patients with PETs. The aim of our study was to determine the sensitivity of EUS-FNA for diagnosing PETs and classifying their underlying malignant potential.Material and methodsWe conducted a retrospective analysis of nonrandomized data of 86 patients, 44 men and 42 women, with an average age of 58 ± 14 years. The tumors were located in the head (52%) in 45 patients, in the body (30%) in 6, in the tail of the pancreas (18%) in 15, with an average size of 29 ± 20 mm(5–100 mm). In 14 patients (16%), the size of the tumor was less than 10 mm, 8 (9%) had a cystic component, 12 (14%) had functional tumors (F-PET). The EUS procedures were performed in a single institution; we used linear-array echoendoscope (FG36X or EG38UT Pentax Europe Ltd., Hamburg, Germany) with an ultrasound platform (Hitachi 6500 or 8500, Hitachi Medical Systems GmbH, Wiesbaden, Germany). FNA was performed by using a 22-gauge FNA needle (Echotip, Cook Endoscopy, Wiston-Salem, NC). There was no on-site evaluation by a cytopathologist. The diagnosis of PET was established on samples obtained by EUS-FNA, and microbiopsies or monolayer cytology (Thinprep®) and/or histopathologic evaluation were performed on surgical specimen of pancreatic tumors or metastases. In each case, antibodies against — synaptophysin, chromogranin A, cytokeratin AE1-AE3, p53, Ki-67 — were used for diagnosis and assessing differentiation. We used the WHO classification [1], which distinguishes three groups of tumors as follows: well-differentiated endocrine tumors, well-differentiated endocrine carcinoma, and poorly differentiated endocrine carcinoma. The classification was based on tumor size, cytology, and percentage of Ki-67 positive cells. In the first group of endocrine tumors, there were two subgroups, with benign behavior and with uncertain behavior. In 2006, there was a proposal of TNM staging and grading [2]. The T depends on tumor size and the grading depends onmitotic count and Ki-67 index.ResultsNinety percent of patients (77 of 86) were diagnosed with PETs obtained by EUS-FNA; 78% were in the first group (67 of 86), 9.3% in the second group (8 of 86), and 2.3% in the third group (2 of 86). 10.4% had inconclusive results by EUS-FNA (9 of 86). The diagnosis was done on surgical specimen of the tumors or metastasis. The overall sensitivity rate was 89.6%: 83% for functional PETs versus 91% for nonfunctional PETs, 86% for tumors less than 10 mm versus 88% for tumors more than 10 mm, 90% for solid tumors versus 88% for cystic tumors [3], 89% for tumors located in the head, 88% for tumors located in the body, and 93% for tumors located in the tail of the pancreas, without any statistical difference. We classified the tumors according to the WHO classification (Table 1) and the TNM staging (Table 2), and we compared the results of both classifications (Table 3). We found a good correlation between the WHO classification and TNM staging obtained by EUS-FNA. Major discrepancies were found in the group of endocrine tumors of uncertain behavior. We have the follow-up data of 60 patients, 16 patients died during a mean follow-up of 43 ± 27 months. According to the WHO classification, overall 5-year survival was 60% (Table 4). The percentage of survivors decreased depending on the grade. For the inconclusive FNA, the 5-year survival is between groups 2 and 3.CommentsThe overall sensitivity rate of EUS-FNA for diagnosing PETs was 90%. The sensitivity rate did not vary with the size, type, location of the tumors, and the presence of hormonal secretion. There was a good correlation between the WHO classification and the TNM staging obtained by EUS-FNA. This single-center experience demonstrated the accuracy of EUS-FNA in diagnosing and determining the potential malignant behavior of PETs [4,5]. EUS-FNA findings predict 5-year survival in patients with PETs [5]. This might be useful for directing the management of patients with PETs. As in all retrospective studies, our study has some limitations such as underreporting and missing data. Because ours is a reference center, we probably have a selection bias of more advanced disease. Moreover, the mitotic count is very difficult on EUS-FNA specimens, and the Ki-67 index is generally done on a small number of cells less than 2000, recommended in the TNM grading. The size of the EUS-FNA specimens could be also nonrepresentative of the tumor heterogeneity.
English Else