Indications, preparation, and adverse effects

Diagnostic imaging

Endosonographic findings can be diagnostic for certain lesions, including gut duplication cysts, lipomas, bile duct stones, and some branch duct intraductal papillary mucinous neoplasias. However, in other situations, EUS imaging alone does not provide a confident diagnosis, and EUS-guided FNA or core biopsy (FNB) may be indicated to facilitate diagnosis. Follow-up imaging may be indicated when EUS demonstrates a benign-appearing lesion—to identify interval growth or other signs suggestive of malignancy.

Tumor staging

Initial evaluation of patients with gastrointestinal (GI) cancers includes assessment of operative risk and determination of tumor stage. Accurate staging determines prognosis and guides treatment decisions. Staging usually begins with noninvasive imaging such as computed tomography (CT), magnetic resonance imaging (MRI), or positron emission tomography (PET), which are generally superior to EUS for identifying distant metastases. EUS is often performed for T staging and nodal (N) staging because it provides an accuracy of approximately 85% for locoregional staging of GI luminal cancers. T-staging accuracy may be lower in very large tumors, and distinguishing T1a from T1b epithelial neoplasms is challenging with standard echoendoscopes. One recent study reported that submucosal saline injection significantly improved the diagnostic accuracy of EUS in distinguishing between T1a and T1b esophageal squamous cell carcinoma, improving triage of patients to endoscopic versus surgical resection. Prior radiation therapy substantially decreases the T-staging accuracy of EUS. Findings of EUS and CT or MR are complementary for T staging in some situations; for instance, staging of vascular involvement by pancreatic cancer.

EUS provides important N-staging information in patients with lung, esophageal, and rectal cancer. The number of malignant lymph nodes and their location may both have staging and prognostic relevance. The typical EUS characteristics of malignant lymph nodes are hypoechogenicity, round shape, smooth border, and size greater than 1 cm in the short axis; however, these features are at best 75% accurate for predicting malignancy in lymph nodes when compared with EUS FNA results or surgical histology. Overlap in appearance between benign and malignant lymph nodes makes N staging problematic by EUS, and the aforementioned criteria were initially validated in esophageal cancer patients and are less useful in lung cancer, rectal cancer, and cholangiocarcinoma Overstaging may result from enlarged reactive lymph nodes that are deemed malignant on the basis of their EUS appearance alone. The addition of FNA improves N-staging accuracy, but it also introduces the possibility of false-positive results, particularly when luminal cancer or Barrett’s esophagus is present. When aspirating lymph nodes, avoid traversing the primary tumor to minimize the risk of a false-positive cytologic finding and tumor seeding.

Although EUS has a limited role in establishing the presence or absence of distant metastasis (M stage), studies have reported a high diagnostic accuracy of EUS-guided paracentesis and aspiration of omental deposits in the diagnosis of peritoneal carcinomatosis. ^, Sometimes a suspicious lesion (such as a left adrenal mass in a patient with lung cancer) is best approached for tissue sampling via EUS, or a previously unsuspected metastasis is diagnosed during EUS performed for local staging (e.g., a left lobe liver lesion in a patient with pancreatic cancer). EUS FNA appears both accurate and reasonably safe when sampling the liver and adrenal glands. ^, Vascular cuffing by tumor or distant perivascular metastases can also be detected by EUS and confirmed by EUS FNA.

EUS has been compared with PET in staging of esophageal and lung cancers. PET has the ability to identify distant metastatic disease more accurately than EUS and CT, upstaging patients who were previously considered to have localized disease and excluding the possibility of curative R0 surgical resection. However, PET has limited accuracy in staging local and regional diseases, and EUS remains superior to PET or CT for this indication, changing staging or altering the extent of the radiation field. It appears that PET and EUS are complementary for optimal staging. ^,

In patients with a pancreatic mass that is visible on CT, EUS and CT provide comparable accuracy with regard to vascular invasion and nodal involvement. However, EUS retains a key role in the evaluation of suspected pancreatic cancer for two reasons: EUS can detect abnormalities missed by CT and provides a preferred means of obtaining tissue specimens during the examination. EUS can identify both primary and metastatic lesions that were not identified on CT, including pancreatic masses and left lobe liver metastases, perivascular cuffing by tumor, and malignant involvement of celiac ganglia. The ability to obtain tissue specimens from these sites or from the primary pancreatic mass is important both for diagnosis and staging. Pancreatic mass lesions may be adenocarcinomas, other neoplasms such as solid and cystic papillary neoplasms, neuroendocrine tumors or metastases, or benign conditions such as autoimmune pancreatitis. These lesions cannot always be differentiated by clinical findings, imaging, and laboratory tests, and EUS FNA and FNB allow efficient diagnosis in many such cases. Because EUS is superior to CT for detection of small pancreatic cancers, it should be performed if clinical or CT findings raise the question of a pancreatic tumor not visualized by CT.

EUS has a role in staging non-small cell lung cancer (NSCLC) because CT and PET have poor accuracy for the detection of mediastinal lymph node metastases. Lung cancer patients without suspicious mediastinal adenopathy on CT have up to a 35% prevalence of malignant mediastinal adenopathy. To limit false-positive and false-negative diagnoses of nodal stage, lymph node tissue sampling is advocated when it changes the management strategy (typically when a visualized lymph node is contralateral to the primary tumor). Sampling of all relevant nodal stations traditionally required surgical mediastinoscopy, however a combination of EUS and endobronchial ultrasound (EBUS) for staging in NSCLC has a negative predictive value of 97% in the evaluation of mediastinal lymph nodes. A strategy of EBUS and EUS, with FNA as appropriate, is more accurate and cost-effective than mediastinoscopy. EUS and EBUS are complementary, because neither test visualizes all relevant mediastinal lymph node stations. Both examinations can be performed under the same sedation and using the same EBUS scope. EUS also allows evaluation of the left adrenal gland for previously undetected distant metastases.

Tissue acquisition

The development of linear EUS technology in the early 1990s allowed for EUS FNA and FNB of lesions within and extrinsic to the GI tract wall. Common indications for FNA include pancreatic mass lesions and N staging of esophageal, pancreatic, bile duct, and rectal cancers. EUS FNA is indicated when it is the least invasive and/or most successful means of obtaining tissue specimens. Molecular characterization of tissue or cyst fluid acquired during EUS is increasingly relevant to clinical care because it may determine diagnosis, predict prognosis, and guide individualized patient management.

Less invasive approaches for establishing a tissue diagnosis include transabdominal ultrasound or CT-guided biopsy. These methods are well established and more cost-effective than EUS in certain scenarios (e.g., in patients with a pancreatic mass and suspicious liver lesions that can be targeted for percutaneous biopsy). However, these methods may be limited by their poor sensitivity in the diagnosis of small lesions or by concern for potential tumor seeding of the biopsy needle tract. EUS may be favored in these situations, as well as when EUS is indicated for other reasons, such as locoregional staging or celiac plexus neurolysis. Although the diagnostic accuracy of EUS FNA for liver and nodal metastases is generally greater than 85%, this method is less accurate in other settings, including diagnosis of pancreatic cystic lesions, stromal tumors, previously radiated lesions, and autoimmune pancreatitis. EUS FNB safely improves the diagnostic accuracy of EUS in selected settings.

Therapy

EUS facilitates access to target structures for therapeutic intervention. EUS may guide direct placement of lumen-apposing stents across the gut wall, and an EUS needle can serve as a conduit that allows for the passage of guidewires or placement of materials with therapeutic intent. The first such therapies to be developed were EUS-guided celiac plexus/ganglia neurolysis or block ^, and EUS-guided pseudocyst drainage. EUS-guided drainage has become a standard of care for patients with walled-off pancreatic necrosis and other pancreatic fluid collections requiring drainage. EUS is also used to facilitate drainage of otherwise inaccessible biliary and pancreatic ducts, angiotherapy for GI bleeding, placement of fiducials to guide radiation therapy, recovery of migrated stents, transduodenal gallbladder drainage, abscess drainage, and creation of gastrojejunostomies and other enteroenteric anastomoses. EUS fine-needle injection (EUS FNI) is a developing means of delivering therapeutic agents into solid cancers and cystic neoplasms. The safety, efficacy, and clinical role of these procedures are discussed in more detail in other chapters.

Contraindications

Absolute contraindications to EUS are few and include unacceptable sedation risks, known or suspected acute viscus perforation, and patient refusal to give consent. Coagulopathy (international normalized ratio [INR] >1.5), thrombocytopenia (platelets <50,000), and ongoing use of antithrombotic agents are relative contraindications to EUS FNA, which should be deferred if possible while coagulopathy or thrombocytopenia are corrected; in some circumstances, detailed later in this chapter, it may be reasonable to proceed with EUS FNA without discontinuing antithrombotic agents. Other relative contraindications include newly diagnosed cancer in a patient who has not undergone appropriate initial evaluation, altered anatomy prohibiting access via EUS, and the presence of intervening vital structures in the planned needle path for EUS FNA.