Endoscopic imaging and image-guided sampling for pancreatic neoplasia

Background: Endoscopic ultrasound (EUS) provides high-resolution images of the pancreas. There are two fundamental echoendoscope designs, yet no data support use of one type over the other. Once lesion is detected, EUS-guided fine-needle aspiration (FNA) is performed, but standard (STN) processing techniques can be associated with limited cellularity, leading to evaluation of novel filter-clot (FC) cellblock technique. With improved quality of EUS-FNA samples, it may be possible to pre-operatively evaluate biomarker status of pancreatic lesions to better risk stratify and tailor future treatments to individual patients. Aims: 1) To compare pancreatic lesion detection rates using radial and linear EUS and to determine incremental diagnostic yield of second EUS exam in tandem study. 2) To compare diagnostic yield and accuracy of STN and FC techniques of processing EUS-FNA samples. 3) To determine whether DPC4 gene status using EUS-FNA samples correlate with clinical outcomes in pancreatic cancer (PC) patients. Methods: High-risk individuals (HRIs) in screening program underwent radial or linear EUS or tandem radial and linear EUS in randomized order. Pancreatic lesion detection rates were compared. EUS-FNA samples during a 13-month period were used to compare the sample adequacy and diagnostic accuracy of STN and FC techniques. Retrospective review was performed evaluating whether Dpc4 immunolabeling status in EUS-FNA samples of PC patients correlated with pattern of disease progression and clinical outcomes. Results: In HRIs, linear EUS detected more pancreatic lesions than radial EUS. In tandem EUS exams, second EUS detected additional lesions, whether the initial exam was performed with radial or linear EUS. However, incremental detection rate was significantly higher if the second exam was linear EUS. FC technique had higher sample adequacy and diagnostic accuracy than STN technique. DPC4 status was associated with pattern of failure, with DPC4 negative PC more likely to progress with metastasis. Conclusion: Linear EUS should be the echoendoscope of choice in pancreatic imaging. FC technique was associated with improved quality of EUS-FNA samples. DPC4 status correlated with disease progression phenotypes in PC patients. Pre-operative biomarker studies using EUS-FNA samples have the potential to advance the concept of personalized medicine in the field of pancreatic oncology