Lung cancer is the leading overall cause of cancer death in the United States. Metastasis to the mediastinal lymph nodes occurs in nearly one-half of patients. Accurate staging of patients with non-small-cell lung cancer (NSCLC) is critical for prognosis, selection of those with potentially resectable and curable disease, and determination of eligibility for clinical research studies. With the improvements in image resolution and the increasingly complex use of multimodal therapies, more patients with suspected lung cancer require precise staging. Pathologic confirmation of clinically suspicious mediastinal lymph nodes is required to select patients for surgery with curative intent.
Although an increasing variety of competitive and complementary staging techniques are available, there is no consensus on how best to stage patients with the greatest accuracy and least morbidity. Endoscopic ultrasound (EUS) with fine-needle aspiration (FNA) has emerged over the past decade as a minimally invasive and effective technique to evaluate the posterior mediastinum. The ideal staging approach is non- or minimally invasive, accurate, safe, and cost-effective. Rapidly evolving techniques such as integrated 18F-fluorodeoxyglucose positron-emission tomography (PET)/computed tomography (CT), and endobronchial ultrasound (EBUS) need to be assessed in the context of well-established techniques such as transbronchial fine-needle aspiration (TBNA) and mediastinoscopy.
This review will focus on the comparative performance and future applications of EUS-FNA in staging patients with suspected or known NSCLC.
Patients with ipsilateral or subcarinal mediastinal lymph node metastases (N2) or contralateral mediastinal lymph node involvement (N3, stage IIIB) are not appropriate for surgical resection. Patients with regional metastases are best approached with multimodality therapy, while surgery alone is reserved for patients without nodal and/or distant metastases. The vast majority of patients with suspected NSCLC undergo initial staging with CT of the chest, or increasingly, CT with integrated PET-CT. These examinations, however, are not recognized as definitive proof of N2-N3 disease. Furthermore, while CT can in some circumstances identify unresectability (invasion of the great vessels), it lacks sufficient sensitivity to detect small mediastinal metastases.
Toloza et al conducted a meta-analysis of 20 studies with a total of 3,438 patients and reported a sensitivity of 57% and specificity of 82% for CT in the detection of positive mediastinal lymph nodes. While PET is widely thought to be more accurate than CT, false-positive results are common (up to 39%). PET alone cannot generally differentiate left- from right-sided hilar activity. Toloza et al performed another meta-analysis of 18 studies with 1,045 patients and reported a pooled sensitivity of 84%, specificity of 89%, positive predictive value (PPV) of 79%, and negative predictive value (NPV) of 93% for PET scanning in staging mediastinal lymph nodes in NSCLC patients.
TBNA is used to assess the posterior mediastinum. However, it is an invasive, blind technique associated with complications such as bleeding and pneumothorax. Mediastinoscopy—long considered the gold-standard in this setting—is the most invasive technique. It is costly, requires general anesthesia, may necessitate hospital admission, and while safe, is associated with the greatest morbidity.[10,11]
EUS is an outpatient, real-time guided technique that has proven safe and effective in the evaluation of the posterior mediastinum. Additionally, EUS can detect celiac, left adrenal, and some hepatic metastases. EUS does not, in most cases, visualize the anterior mediastinum and, as a highly specialized technique, has been slow to be adopted as part of the routine work-up of patients with NSCLC.
EUS has been used in the diagnosis and staging of gastrointestinal malignancies for more than 20 years. Using a transesophageal approach, EUS can evaluate the posterior mediastinum (subcarina), the aortopulmomary window, left paratracheal area, and select distant sites such as the celiac and left adrenal axis. Initially, EUS was performed without biopsy as an imaging test to evaluate mediastinal lymph node metastases. Several sonographic features of malignant lymph nodes have been proposed-features such as size > 1 cm, rounded, hypoechoic, and sharply defined. These sonographic criteria however do not reliably distinguish malignant nodes from benign reactive lymph nodes. The sensitivity and specificity of EUS without FNA for diagnosing mediastinal lymph node metastases ranges from 54% to 75%, and 71% to 98%, respectively.[12,13] The introduction of FNA for tissue confirmation has markedly improved accuracy to 94%-95%.[14-16]
EUS-FNA can be performed in an outpatient setting. The technique, procedure time, and complication rate is similar to that of standard esophagogastroduodenoscopy (EGD). FNA is performed under direct visualization using a flexible linear-array echoendoscope (Figures 1 and 2). The surrounding blood vessels can be visualized by Doppler sonography. Patients can be typically discharged after 30 minutes of postprocedural observation. Reports of bleeding, infection, and pneumothorax are extremely rare. An attendant cytotechnician or cytopathologist increases the yield and accuracy of EUS-FNA.