Endoscopic ultrasound and pancreatic tumors

Key points

Endoscopic ultrasound (EUS) is the most sensitive imaging modality for the diagnosis of pancreatic ductal adenocarcinoma, especially lesions less than 2 cm in size.
EUS is also the most sensitive imaging modality for the detection of small pancreatic neuroendocrine tumors (PNETs) and is superior to both computed tomography (CT) and magnetic resonance imaging (MRI).
CT is the most accurate modality for the determination of tumor staging and assessment of resectability with exception of detecting portal vein invasion and small-volume ascites, where EUS may be superior. Detection of small hepatic metastases may also be more accurate with MRI and EUS.
EUS-guided fine-needle aspiration/biopsy (EUS FNA/FNB) is the gold standard for tissue acquisition in pancreatic mass lesions. Diagnostic yield can be optimized by the use of second-generation core biopsy needles and sampling lesions adopting the fanning maneuver.

Examination checklist for evaluation of a suspected pancreatic tumor

Tumor

The following characteristics of all visualized masses should be noted: maximal dimensions, irregular or well-defined borders, echogenicity, associated cystic structures, and the presence of pancreatic duct cutoff or dilation.

Vascular invasion

For tumors in the pancreatic head, their relationship to the portal vein, portosplenic confluence, superior mesenteric vessels, hepatic artery, and gastroduodenal artery should be noted. For tumors in the pancreatic body, their relationship to the celiac artery, superior mesenteric artery (SMA), portal confluence, hepatic artery, and splenic vessels should be defined. For tumors in the pancreatic tail, the splenic vessels should also be interrogated. The relationship between the tumor and the vessels should be carefully examined. Notation may be stated as follows: intact hyperechoic tumor/vessel interface, adherent to vessel wall without irregular interface, irregular tumor/vessel interface, tumor invasion, or occlusion of the vessel. For occlusion of the portal or superior mesenteric vein (SMV), venous collaterals in the liver hilum or periduodenal region may be seen. For splenic vein occlusion, collaterals in the splenic hilum or gastric fundus may occur.

Lymph nodes

The following stations should be examined for possible metastatic disease: celiac axis, peripancreatic region (including head, body, and tail), porta hepatis, gastrohepatic ligament, aortocaval, and possibly posterior mediastinal stations. Metastatic lymph nodes will usually be round, well defined, hypoechoic, and at least 5 mm in diameter. However, not all malignant lymph nodes will have all of these features. If a suspected lymph node is identified, its characteristics and distance from the tumor should be noted. EUS FNA/FNB should be performed on suspected distant metastatic lymph nodes.

Liver

Transgastric and limited transduodenal examination of the liver should be examined for metastatic lesions. Liver metastases from primary pancreatic cancer are usually hypoechoic and well defined. One or more than one lesion may be identified. EUS FNA of any suspected lesion should be performed when accessible.

Ascites

This usually appears as a triangular or irregularly shaped anechoic region just outside the duodenal or gastric wall. It may be seen from peritoneal metastases or chronic venous occlusion. Omental nodules may also be visualized. EUS-guided fluid aspiration or biopsy of a nodule should be performed when possible.

Endoscopic ultrasound-guided tissue acquisition

Tissue sampling should be performed from the most distant metastatic site first. If ascites, a distant metastatic lymph node, omental nodule, or a suspicious liver lesion is present, one of these lesions should be sampled first. If these test negative for malignancy, either the suspected tumor or a regional lymph node may be sampled. The following information should be noted from each site biopsied: number of passes performed, route of tissue acquisition, whether suction was used, and results of onsite evaluation.

Staging

All suspected malignant tumors of the pancreas should be identified in terms of tumor-node-metastasis (TNM) staging based on the most current American Joint Committee on Cancer (AJCC) staging classification.

Pancreatic ductal adenocarcinoma

Background

Pancreatic ductal adenocarcinoma is the fourth leading cause of cancer-related death in the United States, with 48,000 deaths per year and a dismal overall 5-year relative survival rate of only 10%. Early diagnosis and management planning are therefore essential for optimal clinical outcomes in these patients, as survival is largely dependent on cancer stage, with 5-year survival rate of 39% for localized disease compared with only 3% for metastatic disease. EUS combined with FNA/FNB is the most accurate modality available for the diagnosis of pancreatic cancer and has thus resulted in a paradigm shift in the management of these patients.

Detection of pancreatic ductal adenocarcinoma

Endoscopic ultrasound versus computed tomography

EUS is a very sensitive modality for detecting pancreatic cancer, with a sensitivity of 89% to 100%, specificity of 50% to 100%, and accuracy of 94% to 96%. EUS has a high negative predictive value of 100%, with two studies showing that when patients with suspected pancreatic cancer had a normal EUS examination, pancreatic cancer was not diagnosed in any of them on long-term follow-up. ^, Therefore a normal-appearing pancreas on EUS examination can reliably exclude pancreatic cancer.

Pancreatic cancer is seen as a hypoatteunating mass on CT in the majority of cases, although it can be isoattenuating in 10% of patients, in which case it is more difficult to visualize. The optimal type of CT imaging in patients with suspected pancreatic cancer is a contrast-enhanced, multiphasic (arterial phase to assess arterial involvement, pancreatic phase to visualize the mass, and portal venous phase to assess involvement of the portal vein and SMV) multidetector CT with minimal slice size of 5 mm, which has a sensitivity of 70% to 100%. Helical CTs appear to have lower sensitivity (63% to 77%) than multidetector CTs for detecting pancreatic tumors. ^, ^, ^, Also, the diagnostic sensitivity of CTs for detecting small pancreatic tumors less than 2 cm in size is only 77%.

Studies comparing CT with EUS for the detection of pancreatic cancer have shown equivocal results. Although older studies demonstrate superiority of EUS over helical CT, ^, more recent studies show EUS to be either marginally superior ^, ^, ^, or only comparable to multidetector CTs. Nevertheless, EUS appears to have a clear advantage over CT in three specific scenarios. First, studies have shown that EUS is superior to multidetector CT for the identification of small pancreatic mass lesions less than 2 cm in size ( Fig. 13.1 A – C). This is probably because small pancreatic mass lesions are more likely to be isoattenuating rather than hypoatteunating on CT and hence must be diagnosed by identifying secondary signs of pancreatic cancer, such as cutoff of the pancreatic duct, pancreatic duct dilation, atrophy of the pancreatic parenchyma, or absence of normal pancreatic contour. Second, EUS can diagnose pancreatic cancer in patients without a distinct mass lesion on CT but with a high clinical suspicion with a sensitivity of 68% to 97%, specificity of 43% to 71%, and accuracy of 72% to 92%. ^, Three, EUS has been shown to be particularly useful in detecting pancreatic cancer that was missed on CT or MRI in 1.25% to 5% of patients with unexplained acute or chronic pancreatitis. ^, Therefore EUS should be performed in CT-negative patients in whom the clinical suspicion for pancreatic cancer remains high.

• Fig. 13.1, Computed tomography of a patient who presented with obstructive jaundice but no definitive mass on imaging (A). Linear endoscopic ultrasound imaging revealed a 2-cm pancreatic head mass (B) that on rapid onsite evaluation (ROSE) was proven to be an adenocarcinoma (C) (Diff-Quik staining 200×). CBD , common bile duct; PV , portal vein.

Endoscopic ultrasound versus magnetic resonance imaging

MRI can provide detailed imaging of the pancreatic parenchyma and pancreatic duct and has a sensitivity of 78% to 100% and an accuracy of 79% to 91% ^, ^, ^, ^, for detecting pancreatic tumors. In studies that compared EUS with MRI for the diagnosis of pancreatic cancer, EUS was at least equivalent or superior to MRI for visualizing pancreatic cancer. ^, In one study of 62 patients, MRI had the lowest accuracy for tumor detection at 62% when compared with both EUS (63%) and CT (73%).

Contrast-enhanced endoscopic ultrasound

Pancreatic adenocarcinoma is visualized as a hypoenhancing lesion on CE EUS, whereas chronic pancreatitis is either isoenhancing or hyperenhancing. In a meta-analysis of 1139 patients, contrast-enhanced EUS was shown to have a high sensitivity for diagnosing pancreatic cancer, with a pooled sensitivity of 94% and a specificity of 89%. When compared with conventional EUS for performing FNA, although there was no significant difference in diagnostic performance between modalities, fewer passes were required to obtain a diagnostically adequate sample with contrast enhancement. However, in a recent randomized trial that compared contrast-enhanced versus conventional EUS-guided FNA/FNB of pancreatic masses, there was no difference in diagnostic sensitivity between groups. Therefore, while contrast-enhanced harmonic EUS may not have a supportive role in routine EUS FNA/FNB evaluation of all pancreatic mass lesions, they may be selectively considered for identifying and sampling small or indeterminate lesions, especially in patients with underlying chronic pancreatitis.

Endoscopic ultrasound elastography

EUS elastography utilizes the difference in tissue elasticity to distinguish between malignant and benign lesions. EUS elastography can produce either a qualitative assessment with generation of different color hues or a quantitative assessment with strain ratios. Several meta-analyses have been published on this imaging technique, with a sensitivity of 95% to 98% and a specificity of 67% to 76% for diagnosing pancreatic cancer.

Assessment of resectability and cancer staging

Surgical resection is the only currently available treatment with potential for cure in patients with pancreatic cancer. Even though the 5-year survival remains poor at 10% to 20% postresection, ^, survival has been shown to be significantly longer in patients undergoing margin-negative (R0) resection compared with nonsurgical candidates (14 vs 5 months, P < .0001). Therefore accurate and timely pancreatic cancer diagnosis and staging are paramount in order to expedite referral to surgery.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here