The Role of EUS in Diagnosis of Pancreatic Cancer

EUS has an integral role in the diagnosis of pancreatic cancer given its high sensitivity for detecting pancreatic neoplasms and the access it affords to perform fine needle aspiration (FNA) of the suspected lesions. The sensitivity of EUS for detecting pancreatic lesions ranges from 85% to 99% in most published series. The sensitivity and accuracy of EUS are slightly higher than the sensitivity and accuracy of computed tomography (CT) in detecting small pancreatic lesions. Magnetic resonance imaging (MRI) has similar accuracy to CT scanning in detecting pancreatic cancer and vascular invasion.

Pancreatic lesions usually appear on EUS as hypoechoic lesions with an irregular border. TNM staging of pancreatic tumors by EUS is feasible and accurate. If the tumor is limited to the pancreas, it is either a T1 or T2 lesion. Lesions smaller than 2 cm are T1; lesions larger than 2 cm are T2. If the tumor extends beyond the pancreas, then it is either a T3 or T4 lesion. Tumors extending to the celiac artery or superior mesenteric artery are considered T4 lesions, whereas tumors involving any of the surrounding structures of the pancreas, such as the portal vein, duodenum, or ampulla of Vater, without involvement of the celiac artery or superior mesenteric artery are classified as T3 tumors. The distinction of a T3 from a T4 tumor is important given that tumors extending to the celiac or superior mesenteric arteries (T4) are generally not surgically resectable for cure. If malignant lymph nodes are seen around the pancreas, such as the peripancreatic, celiac, or gastrohepatic lymph nodes, then the lesion is an N1 lesion. Occasionally, distant metastasis can be seen by EUS and, in this case, the lesion is an M1 lesion, which is also inoperable. Soriano and colleageus compared the accuracy of CT, MRI, and EUS in assessing TNM staging of pancreatic cancer using surgical diagnosis as the gold standard. CT had the highest accuracy for T-staging at 74%; MRI accuracy was 68% and EUS, 62%. However, EUS had the highest accuracy for N-staging (65%) with accuracy of CT and MRI for assessing N-staging at 62% and 61%, respectively. In a prospective study by Soriano and colleagues, CT had the greatest accuracy in detecting vascular invasion (83%) in comparison with EUS and MRI. Similar results were seen in another trial by Tian and co-workers, in which they found that helical CT had the highest accuracy in detecting vascular invasion of pancreatic cancer, and EUS had the highest accuracy of assessing lymph node metastases. It is worth stressing that, although CT is more accurate in assessing T-staging of pancreatic cancer, EUS is still more sensitive and accurate in pancreatic lesions less than 3 cm in size.

FNA can be done with a variety of dedicated needles, most typically 22 gauge. Other needles available are the 25-gauge, 19-gauge, and core biopsy needles. The number of passes needed to obtain sufficient tissue for diagnosis is higher in pancreatic cancer than in other lesions, and varies from 4 to 7 passes among the studies. Combining cytologic and histologic analyses of the specimen can decrease the number of passes to 2. There is a growing body of evidence that the 25-gauge needle may be superior to the 22-gauge needle in the diagnosis of pancreatic cancer. Perhaps this is because less blood is aspirated with the 25-gauge needle, improving the cytologic assessment of the specimen. In addition, the 25-gauge needle is easier to use in areas with tough angulation, as in the case of lesions sampled from a duodenal sweep. On-site cytologic interpretation has been shown in many studies to improve yield and reduce the number of passes needed.