Endoscopy 2020; 52(S 01): S318
DOI: 10.1055/s-0040-1705025
ESGE Days 2020 ePoster presentations
Thursday, April 23, 2020 09:00 – 17:00 Endoscopic ultrasound ePoster area
© Georg Thieme Verlag KG Stuttgart · New York

WHICH EUS FEATURE BEST PREDICT THE DIAGNOSIS OF PANCREATIC NEUROENDOCRINE TUMORS

A Saizu
1   Clinical Institute Fundeni, Gastroenterology, Bucharest, Romania
,
G Becheanu
1   Clinical Institute Fundeni, Gastroenterology, Bucharest, Romania
,
M Dumbrava
1   Clinical Institute Fundeni, Gastroenterology, Bucharest, Romania
,
R Costache
1   Clinical Institute Fundeni, Gastroenterology, Bucharest, Romania
,
R Iacob
1   Clinical Institute Fundeni, Gastroenterology, Bucharest, Romania
,
C Gheorghe
1   Clinical Institute Fundeni, Gastroenterology, Bucharest, Romania
› Institutsangaben
Weitere Informationen

Publikationsverlauf

Publikationsdatum:
23. April 2020 (online)

Auch verfügbar auf
 

Aims Endoscopic ultrasound (EUS) associated with fine needle aspiration (FNA) has become the standard when it comes to evaluate pancreatic mass lesions. Our aim is to determine if beside specific ultrasound characteristics, a particular hypoechoic rim feature can predict a pancreatic neuroendocrine tumor (pNET) diagnosis.

Methods We have conducted a retrospective analysis of EUS examinations performed for solid pancreatic lesions from 2015 to 2019 in a Tertiary Gastroenterology Center. There were 175 cases analyzed, out of which pNET diagnosis was established by FNA in n= 19 (10.8%). Mean age of patients with pNETs was 56.1 ± 10.3 years, significantly lower than all cases in the study group and 47.3% of them were males. The following EUS features were analyzed by univariate and multivariate analysis: well defined margins, hypoechoic EUS aspect, homogeneous pattern, the presence of internal vascularization, stiff elastographic appearance, the presence of multiple lesions, the internalization of large vessels, Wirsung duct dilation and the presence of a hypoechoic rim delimiting the interior margin of the lesion.

Results The following EUS features were significantly associated with EUS diagnosis of pNET: well defined margins (63.1% vs 35.7% p=0.02), homogeneous pattern (36.8% vs 16.9%, p=0.03), the presence of internal vascularization (63.1% vs 24.3%, p=0.0004), blue on elastography (89.4% vs 64.2%, p=0.03) and the presence of a hypoechoic rim (47.3% vs 2.5%, P< 0.0001). In the multivariate logistic regression analysis only the presence of a hypoechoic rim was independently associated with the pNET diagnosis (p< 0.0001).

Conclusions EUS appearance can suggest the diagnosis of a pNET. The presence of a hypoechoic rim delimiting the interior margin of the lesion is an independent predictor of pNET diagnosis. During EUS for solid pancreatic masses, the presence of this rim, should determine the endoscopists to request further immunohistochemical stains.