الفهرس 
IntroductionOnly 14 pages are availabe for public view
 |  | from | 163 |  |  |
| | | from | 163 | | |
Abstract
Accumulated evidence has revealed that endoscopic ultrasonography
(EUS) has had a great impact on the clinical evaluation of pancreatic
tumors.
Accordingly, the aim of this work was to assess the role of EUS in the
diagnosis of pancreatic tumors.
This was an observational cross-sectional study conducted at the
Alexandria Fever Hospital from January 2018 to December 2021. The
study included 80 patients who presented with clinical and laboratory
manifestations of pancreatic tumors and were subjected to EUS.
Most of the patients were males (62.5%), with ages ranging from 28
to 83 years with a mean of 56.6 ± 12.52 years. Most of them reported a
history of jaundice (72.5%) and abdominal pain (75.0%). Only 27.5% of
them reported having palpable abdominal mass on examination.
The commonest site of pancreatic tumors was the head of the
pancreas (55%) followed by the pancreatic body (17.5%), ampulla
(10.0%), the uncinate process (7.5%), and lastly, only 5% had tumor site at
the CBD and Tail. Also, most of the cases (62.5%) were with a size ≤3 cm.
Using EUS, most of the lesions were solid (60.0%), had ill-defined
margins (50%), were heterogeneous (47.5%), and hypovascular (75.0%).
Extra-pancreatic extension was detected in 35%. Also, 47.5% had
associated lymph nodes while only 30% had vascular involvement. Dilated
CBD was demonstrated in 65% and dilated pancreatic duct in 57.7%.
Regarding the consistency during the FNA, the majority of them (62.5%)
were firm while 27.5% were soft and only 10% were hard.
As regards CT, 45.0% of cases had a malignant lesion, 37.5% had
benign lesions and 17.5% had an inflammatory lesion. By EUS, there were
55.0% with malignant lesions, 27.5% with benign lesions, and 17.5% with inflammatory lesions. Finally, regarding the histopathology, 57.5% of
cases had a malignant lesion, 30.0% had benign lesions and only 12.5%
had inflammatory lesions.
The findings of the EUS and CT in comparison to the histopathological diagnosis as the gold standard for identification of the lesion
revealed an accuracy for diagnosis of 90% and 70% for EUS and CT
respectively.
There was an excellent diagnostic performance of EUS in diagnosing
malignant, benign, and inflammatory pancreatic lesions. The accuracy of
EUS for diagnosing malignant lesions was 93% with 91% sensitivity, 94%
specificity, 95% PPV, and 89% NPV. For benign lesions, EUS showed an
accuracy of 93% with 83% sensitivity, 96% specificity, 91% PPV, and
93% NPV. While the EUS accuracy was 95% with 100% sensitivity, 94%
specificity, 71% PPV, and 100% NPV for diagnosing inflammatory
pancreatic lesions.
There was a moderate to excellent diagnostic performance of CT in
diagnosing malignant, benign, and inflammatory pancreatic lesions. The
accuracy of CT for diagnosing malignant lesions was 83% with 74%
sensitivity, 94% specificity, 94% PPV, and 73% NPV. For benign lesions,
CT showed an accuracy of 78% with 75% sensitivity, 79% specificity,
60% PPV, and 88% NPV. While the CT accuracy was 80% with 40%
sensitivity, 86% specificity, 29% PPV, and 91% NPV for diagnosing
inflammatory pancreatic lesions.
There was a good agreement between the EUS and histopathology
regarding the diagnosis of malignant, benign, and inflammatory pancreatic
lesions. While the agreement of EUS and CT together and CT and
histopathology together as regards the diagnosis of malignant, benign, and
inflammatory pancreatic lesions was poor to moderate.