Introduction

Transbronchial needle aspiration (TBNA) is a technique that has revolutionized the diagnosis of mediastinal pathology by enabling intrathoracic nodal sampling in a minimally invasive manner. The sampling of paratracheal masses using an esophageal varix needle passed through a rigid bronchoscope was initially described by Ko-Pen Wang in 1978. The following year Oho and colleagues created a needle that could be passed through a flexible bronchoscope, thus ushering in a novel modality for sampling intrathoracic lymph nodes without surgical intervention. Over the last 30 years, the technique of TBNA has been relatively unchanged, although the indications for TBNA have expanded to include sampling of hilar lymph nodes, submucosal disease, visible endobronchial lesions, as well as peripheral nodules. The recent development of advanced imaging such as computed tomography (CT) fluoroscopy, electromagnetic/virtual bronchoscopic navigation, and endobronchial ultrasound (EBUS) now allows real-time visualization of lymph node sampling. Because of these advances, TBNA has emerged as the first line of intrathoracic lymph node sampling for the diagnosis, staging, and prognosis of bronchogenic carcinoma, sarcoidosis, and even infectious diseases.

Bronchogenic carcinoma is the leading cause of cancer death in both men and women in the United States as well as in several other countries. Therapeutic options and prognoses are heavily dependent on accurate staging, and nodal staging is a key component of determining overall clinical stage (Figure 11.1). Non-invasive radiologic staging is suboptimal with sensitivities ranging from 51 to 74 percent. Additionally, because radiologic staging has been demonstrated to differ markedly from pathologic staging, the American Thoracic Society (ATS), European Respiratory Society (ERS), and European Society of Cardiothoracic Surgery all recommend that pathologic evaluation of the mediastinum be obtained in all patients prior to surgical resection of lung cancer. Mediastinoscopy has long been the gold standard for obtaining tissue, with sensitivities and specificities of 80–90 percent and 100 percent, respectively. Comparatively, the overall sensitivity of TBNA is 78 percent. Despite a high specificity (99 percent), TBNA is limited by a higher false-negative rate compared with that of mediastinoscopy (28 percent vs. 11 percent). The yield with TBNA has been associated with tumor cell type (small cell > non-small cell > lymphoma), lymph node size, and lymph node location. EBUS-TBNA has been shown to have a sensitivity and specificity of approximately 94 percent and 100 percent, respectively, and its yield may be independent of lymph node size and location. Nonetheless, if TBNA is nondiagnostic, surgical sampling is still recommended to rule out mediastinal nodal involvement. As a minimally invasive procedure, TBNA offers the significant benefit of fewer complications and has been demonstrated to preclude surgery in as many as 66 percent of patients.

Figure 11.1 Possible treatment options for patients with non-small cell lung cancer.

TBNA can also be used for peripheral parenchymal opacities. For peripheral lung cancer, TBNA has been shown to have a higher diagnostic yield than brushing or transbronchial biopsy. Peripheral TBNA is the exclusive diagnostic modality in up to 35 percent of cases, can obtain diagnostic tissue in up to 76 percent of cases, and significantly increases the yield when combined with other modalities used to sample peripheral opacities such as bronchoalveolar lavage (BAL), brushing, and/or transbronchial biopsy.

The major limitation of TBNA is that it remains an underused technique. Two recent surveys have found that only 12 percent of pulmonologists routinely used conventional TBNA in the evaluation of malignant disease, though EBUS is certainly increasing the overall use of TBNA. This likely results from fears of not obtaining adequate tissue, fears of causing damage to the bronchoscope, as well as inadequate training during fellowship. Among these reasons, lack of training is likely the foremost reason for the underuse of TBNA in clinical practice.

Anatomy

It is crucial that the bronchoscopist become an expert in airway anatomy, and there are several excellent chapters and texts to assist with this process. In addition to the upper airway and the segmental anatomy of the lungs, it is essential to have a comprehensive understanding of the anatomy external to the airway, primarily the intrathoracic vessels and lymph nodes. To maximize the sensitivity of TBNA, Wang proposed a staging system and anatomic guide for needle placement as defined by the CT scan and bronchoscopic imaging. Although extremely useful to the bronchoscopist, this lymph node map has one important limitation in that the nodal stations do not correspond exactly to those used by radiologists and thoracic surgeons. For example, in the system proposed by Wang, the left paratracheal station (4L) is also called the “aortic-pulmonary window.” On the ATS/ERS lymph node map, however, the aortic-pulmonary window actually refers to station five and six nodes, which lie on the lateral side of the ligamentum arteriosum, and are therefore inaccessible by TBNA. As communication with our colleagues is essential, we recommend using the system proposed by the IASLC (Figure 11.2). That way, when the surgeons and radiologists refer to a left paratracheal lymph node (station 4L), everyone will be on the same page.

Figure 11.2 IASLC lymph node map (From: J Thorac Oncol 2009;4:568.)

To improve one’s yield, Wang suggested flipping the right and left CT images to help visualize the anatomic location of the lymph node relative to the airway during the procedure. Virtual bronchoscopy, by making the airway wall translucent and highlighting the target lymph node, may also aid the bronchoscopist by having the visual of the relevant anatomy in the “mind’s eye” during the procedure. The major drawback to these techniques is that the bronchoscopist is still required to mentally transpose the images during the procedure, and remains uncertain as to final needle position. Failure to place the needle directly into the lesion remains the leading cause of a lower yield on biopsy. Among lymph nodes of equivalent size, right paratracheal (4R) and subcarinal (7) stations have been shown to have a higher diagnostic yield than nodes in other stations. As such, these should be the first nodal stations attempted by novice operators.

Equipment

Although there are many different types of TBNA needles available for use, they can be divided into two broad categories: histology (19-gauge) and cytology (22-gauge) needles. Some histology needles have an inner 21-gauge needle and an outer 19-gauge needle, whereas others are a single 19-gauge needle. There has never been, and likely will never be, a study showing that one particular needle is better than another. Therefore, the bronchoscopist should familiarize him- or herself with one particular histology needle and one particular cytology needle.

Schenk and colleagues have shown that the 19-gauge needle provides a higher diagnostic yield as compared with that of the 22-gauge needle when used for lung cancer staging. All transbronchial needles consist of (1) a distal retractable sharp beveled needle, (2) a flexible catheter with a metal hub at the distal end, (3) a proximal control device used to manipulate the needle, and (4) a proximal side port through which suction can be applied.

Technique

When used for the staging of non-small cell lung cancer (NSCLC), it is of utmost importance to avoid false-positive results, which would have the clinical effect of upstaging a patient, and potentially precluding curative surgery. For this reason, TBNA should be the first procedure performed during the bronchoscopy, and the lymph node station that would place the patient at the highest stage should be sampled first. For example, in the case of a patient with a T2 lesion in the left upper lobe, with adenopathy in stations 4L, 4R, and 7, the 4R node should be sampled first, as it would be considered N3, and therefore make the patient stage IIIb, and therefore not a surgical candidate. If the bronchoscopist attempted a BAL, brush, and transbronchial biopsy of the left upper lobe lesion first, the working channel of the bronchoscope could become contaminated and provide a false-positive result for subsequent TBNAs.

The Golden Rule of TBNA is “never advance the catheter in the bronchoscope with the needle out.” A corollary to this rule is that the bronchoscopist is only as good as his or her support staff. It is essential to train your assistant prior to performing the procedure to be sure that he or she knows how to operate the needle. When the bronchoscope is in the desired location, the operator should keep the tip of the bronchoscope midline in the airway and advance the needle assembly through the working channel of the bronchoscope, making sure that the needle is retracted into the metal needle hub. After the catheter tip is visible within the bronchial lumen, the needle is advanced and locked into place. If at any time prior to needle insertion into the airway, the needle tip is not visible, it should immediately be retracted by the assistant. Once the bronchoscope is at the level of the target lymph node, the bronchoscope is flexed up so that the needle tip is as perpendicular to the airway wall as possible. Passage of the needle through the airway wall can occur in one of four methods, as described by Dasgupta and Mehta (Figure 11.3).

Figure 11.3 Common techniques for transbronchial needle aspiration. (From: Dasgupta A, Mehta AC, Wang KP. Transbronchial needle aspiration. Semin Respir Crit Care Med. 1997;18:571–581.)

Specimen Processing and Interpretation

All material from the biopsy should be processed to maximize the yield from the procedure. A sample from a 19-gauge needle can be sent in formalin as a “core biopsy” to surgical pathology or be processed by cytopathology. Samples from a 22-gauge needle, however, should all be processed by cytopathology. It is important to talk to the surgical and cytopathology departments at one’s institution to determine how they prefer their specimens. Some institutions have rapid on-site evaluation (ROSE) and may prefer that the specimen be placed directly on a glass slide for immediate staging, whereas other institutions prefer that the material for cytologic analysis be placed in CytoLyt or another medium. Several studies have shown that the use of ROSE is associated with a reduction in the number of inadequate specimens. This may be less of an issue with the development of real-time imaging provided by EBUS, but we still feel that it is important to review the results from all cases. This serves to enhance one’s technique and provide feedback as to what worked and what did not for a particular needle pass. A major caveat for ROSE is the potential for a false-positive result that leads to premature termination of the procedure. As such, it is crucial that the cytopathologists be experienced in interpreting TBNA and only state that they have an adequate tissue sample when they are confident. Additionally, it is crucial to obtain enough tissue for analysis of genetic tumor markers, a diagnosis of “cancer,” or even “adenocarcinoma” isn’t adequate.

As TBNA does not have a 100 percent negative predictive value, we do prefer the term “nondiagnostic” to “negative.” This is an important concept to explain to the patient as well, as, if the TBNA does not provide an answer, it may be essential to take the next step, perhaps a mediastinoscopy, and obtain more tissue.

Complications

TBNA is perhaps the safest bronchoscopic procedure available, with an overall complication rate of 0.26 percent. There have been case reports of bleeding, infection, pneumothorax, and pneumomediastinum; however, serious adverse effects are much less frequent when compared with those from transbronchial biopsy or surgical biopsy.

Improving the Yield

Clearly, experience counts. In a recent study by Hsu and colleagues, the sensitivity of TBNA increased from 33 to 81 percent over a period of 3.5 years. To gain experience, we recommend a commitment to learning mediastinal anatomy, understanding the technique from a conceptual standpoint, practicing the procedure on either a simulator or other model to gain the psychomotor skills, and then performing the procedure on as many patients with mediastinal and hilar adenopathy as possible. Feedback is essential, and it is crucial to review all cytology/pathology samples and to try to reference each pass with the technique used to obtain the tissue. This is a key benefit that ROSE can provide during the learning curve of TBNA.

The use of 19-gauge histology needles has been shown to increase the yield when compared with the use of cytology (22-gauge) needles, especially for nonmalignant disease. As earlier, we recommend that the operator become familiar with both a histology and a cytology needle. If the diagnosis leading the differential is not malignancy, the histology needle should be used.

The use of CT fluoroscopy (CTF) can be used to provide real-time confirmation of needle location. As opposed to standard fluoroscopy, which is typically used only in two dimensions, CT provides the ability to visualize the target in three dimensions and provides real-time three-dimensional confirmation of the appropriate (Figure 11.4A) and inappropriate (Figure 11.4B) biopsy sites. A study by Goldberg and colleagues analyzed data from 12 patients who underwent CTF-guided TBNA via a “quick-check” technique after having previously undergone non-diagnostic conventional BNA. A diagnosis was made in all patients, with a mean procedure time and number of aspirates similar to historical controls. Additionally, CTF confirmed that only six of the initial 18 (33 percent) needle passes were properly positioned within the target lymph node. With guidance, the rate of subsequent successful passes increased to 62 percent. Despite the fact that the TBNAs were performed by an experienced interventional pulmonologist, 42 percent of the needle insertions were not in the intended target. Six insertions (5.2 percent) were seen to have penetrated great vessels (four in the pulmonary artery, and one each in the left atrium and aorta); however, all of these were without complication.

Figure 11.4 CTF showing (A) needle appropriately placed in station 4R node and (B) needle in aorta.

Ernst’s group went on to describe their results obtained by CTF in 32 patients with hilar and mediastinal adenopathy. As the yield with conventional TBNA is relatively high for nodes in the subcarinal and precarinal stations, patients with adenopathy in these areas were required to have a prior nondiagnostic TBNA to be included in the study. Adequate tissue was obtained in 28 patients (87.5 percent), and a specific diagnosis was made in 22 (68.8 percent). One patient had a false-negative CTF-guided TBNA. The authors conclude that, although CTF is not a substitute for good TBNA technique, it is an extremely helpful tool to improve the yield of TBNA in small and less accessible nodes.

EBUS is another important modality used to help improve accuracy of TBNA. Like CTF, EBUS has also been shown to significantly improve the yield of TBNA. Recently, a bronchoscope with a dedicated ultrasound probe and distinct working channel was developed (Olympus Corporation, Tokyo, Japan), and has the benefit of providing real-time guidance for TBNA of mediastinal and hilar lymph nodes, with excellent results. The sensitivity, specificity, and negative predictive value for EBUS-TBNA are 98.7, 100, and 97 percent, respectively. For lymphoma, EBUS-TBNA has been shown to have sensitivity, specificity, and negative predicative value of 91, 100, and 93 percent, respectively, despite the fact that the currently available needle for use in EBUS-TBNA is 22 gauge. Herth and colleagues have recently shown that, even in the presence of a radiographically normal mediastinum (as defined by no nodes > 1 cm on CT), EBUS identified malignancy in nearly 20 percent of patients, and avoided surgery in 17 percent. The use of EBUS has also been shown to improve subsequent yield when performing conventional TBNA.

Summary

Despite TBNA being available for nearly 30 years, it remains one of the most underused tools of the pulmonologist. TBNA is an extremely safe procedure and should be routinely used to sample mediastinal and hilar lymph nodes, endobronchial lesions, as well as parenchymal opacities. EBUS-TBNA now allows real-time guidance of lymph nodes and may preclude the need for surgery in a significant number of patients. These new technologies, however, do not make one a better bronchoscopist; we therefore recommend dedicated training and continued practice.