Bronchoscopy: Transbronchial needle aspiration

INTRODUCTION — Transbronchial needle aspiration (TBNA) of mediastinal lymph nodes was initially described in 1949 and its application via flexible bronchoscopy was later described in 1981 [1-6]. Since then, its utility in the diagnosis of parenchymal, and mediastinal/hilar lesions has been reported in research publications and confirmed with widespread clinical experience [7-14].

TBNA provides an opportunity to diagnose and stage pulmonary, hilar, and mediastinal lesions (particularly lung cancer), as well as evaluate these areas for non-malignant pathology such as sarcoidosis, in a minimally invasive fashion, even in the absence of endobronchial disease [15-17]. Despite such advantages, TBNA has previously been an underutilized technique [18,19], a trend that may change since trainees are increasingly gaining experience with the procedure [20,21].

The role of TBNA in the diagnosis of hilar/mediastinal, central, and peripheral lesions is reviewed here. In addition, anatomic and technical considerations are discussed. The staging of non-small cell lung cancer is presented separately. (See "Overview of the initial evaluation, diagnosis, and staging of patients with suspected lung cancer" and "Tumor, node, metastasis (TNM) staging system for lung cancer".)

DEFINITION — Transbronchial needle aspiration (TBNA) is a procedure to obtain cellular material using a needle that is passed through the bronchial wall. It is used to obtain tissue from lung or hilar/mediastinal lesions that are in close proximity to the endobronchial tree. Bronchoscopy is used to direct the operator to the target lesion (eg, lung mass or lymph node). The catheter and needle are then passed through the working channel of the bronchoscope, through the bronchial wall and material is aspirated for either cytological, histological, or bacteriological analysis. It can be performed as a “blind” procedure during conventional bronchoscopy (bronchoscopic-TBNA) or under image-guidance using a bronchoscope with endobronchial ultrasound or electromagnetic navigational capability (EBUS-TBNA; EMN-TBNA). The term “conventional” TBNA (c-TBNA) is often used to describe the procedure when not performed with EBUS and is preferred over the term “blind” as the bronchoscopist should have reviewed the patient’s imaging (ie, chest computed tomographic [CT] scan) and is performing the procedure with bronchoscopic visualization. Likewise, TBNA needles can be used to sample endobronchial disease and the term endobronchial needle aspiration (EBNA) is often used in the literature or used interchangeably with TBNA when performed for endobronchial disease.

Details of the technique and the indications for bronchoscopy and EBUS are discussed separately. (See "Endobronchial ultrasound: Indications, contraindications, and complications", section on 'Indications' and "Flexible bronchoscopy in adults: Indications and contraindications", section on 'Indications' and 'Technique' below.)

INDICATIONS — The indications for transbronchial needle aspiration (TBNA) are summarized in the table (table 1). The most common application of TBNA is the diagnosis and staging of lung cancer and the evaluation of sarcoidosis. However, TBNA can also be used to sample benign or malignant endobronchial, submucosal, peribronchial (when extrinsic compression exists), as well as parenchymal lesions. TBNA may be the only diagnostic modality that can be performed in patients in whom mediastinoscopy is contraindicated due to a bleeding diathesis [15,16,22].

TECHNICAL CONSIDERATIONS — There are numerous factors that influence the diagnostic yield of transbronchial needle aspiration (TBNA). They include:

●Likelihood of malignancy

●Location (eg, airway lesion, subcarinal and right paratracheal lymph node), size (bulky), and appearance (eg, spiculated masses) of the lesion or lymph node, as determined by computed tomography (CT)

●Visibility and carinal involvement of the tumor, as determined by bronchoscopy

●Image guidance, in particular endobronchial ultrasound

●Operator/team experience

●Cytopathology/pathology experience with small biopsy specimens

In addition, technical issues can influence diagnostic yield, including the type of needle, the number of attempts and aspirates, and in some cases the presence of an on-site cytologist [23-35].

Essentials of the TBNA procedure (the "six T's"; computed tomography, type of needle, technique, tissue preparation, tissue interpretation, and trained assistant), should be considered every time a TBNA is performed (table 2).

Computed tomography — Chest CT is typically performed prior to bronchoscopy and TBNA in order to evaluate the relationship of the tracheobronchial tree to the surrounding structures, including lymph nodes and blood vessels. This maximizes diagnostic yield and safety [36,37]. Virtual bronchoscopy, a CT-based technique that allows the presentation of computer-generated intraluminal images, may be helpful in planning complex procedures. (See "Flexible bronchoscopy in adults: Overview", section on 'Virtual bronchoscopy'.)

●Location of the lymph nodes – TBNA is able to reach the majority of lymph node stations essential in the staging of lung cancer (ie, 2R, 2L, 4R, 4L, 7, 10R, 10L, 11R, and 11L) [36,38]. The most recent International Association for the Study of Lung Cancer (IASLC) lymph node map is shown in the figure (figure 1).

Conventional TBNA of lymph nodes in the subcarinal (station 7) and right paratracheal regions (station 4R) is generally easier, and associated with a higher diagnostic yield than sampling of left paratracheal lymph nodes (station 4L) [39]. As such, given the drainage pattern of tumors, cTBNA of mediastinal lymph nodes performed for diagnosis and staging of a tumor in the right lung is more sensitive than that performed for a tumor in the left lung [40,41].

The location of the lymph nodes in relation to the tracheobronchial tree is best visualized by imagining the interior of the airway as a clock face, using the carina as the central reference point. The best approach for each station with the bronchoscopist standing behind the patient lying in a supine position is outlined in the table (table 3). To obtain an adequate specimen, the needle must reach the core of the lymph node while avoiding nearby vascular structures (figure 2A-C).

The superior vena cava and the azygos vein are found anteriorly and to the right of the distal third of the trachea. Directly anterior to the trachea, above the level of the primary carina, lie the innominate artery and the aortic arch. The main pulmonary artery divides into the right and left branches within the concavity of the aortic arch. The left pulmonary artery runs anterosuperiorly, in close approximation (within 3 to 5 mm) to the left mainstem bronchus, while the right pulmonary artery lies anterior to the right mainstem bronchus and the origin of the upper lobe bronchus. The esophagus lies in close approximation (within 2 to 3 mm) to the posterior wall of the trachea and the left mainstem bronchus [42]. When cTBNA is performed, these areas should be avoided unless a clearly visualized pathologic process is documented by radiographic evaluation.

●Size of the lesion and lymph nodes – Mediastinal lymph nodes that appear enlarged on a CT scan (>1 cm) have better diagnostic yield from conventional TBNA than those that do not appear enlarged (<1 cm) [41,43,44].

Type of needle — A wide variety of biopsy needles are available, but we recommend familiarizing oneself and the bronchoscopy suite staff with no more than one or two needles for each indication.

All needle systems for transbronchial aspiration consist of [45]:

●A retractable sharp beveled flexible needle

●A flexible catheter

●A proximal control device to manipulate the needle, the stylet, or both, and a side port through which suction can be applied

Only retractable needles should be used, because nonretractable needles will damage the working channel of the bronchoscope. The 20 to 22 gauge needles are usually used to obtain cytology specimens while the larger 19 G needles are needed to obtain a core of tissue for histology [45-47]. It should be noted that larger needles may not increase diagnostic yield, and may contain more blood [48]. For lesions located outside the airway wall, the needle should be at least 10 mm in length. The catheter should be flexible enough to maneuver into peripheral locations, yet stiff enough to exert force to penetrate the thick airway wall to sample central lesions.

For staging of patients with lung cancer, some studies suggest that needle size can affect diagnostic yield while others suggest no impact on this outcome [23-27,35,48,49]. Some of this difference may relate to operator experience, the needle size used in the study, the type of bronchoscope used for biopsy (conventional versus endobronchial ultrasound), and the presence of experienced cytopathologists in the bronchoscopy suite (ie, rapid on-site evaluation [ROSE]). One prospective study suggested a higher diagnostic yield of 19 G needles compared with 22 G needles (85 versus 53 percent) when conventional bronchoscopy was used [23]. In contrast, most other studies suggest that there is no difference in the diagnostic yield with a 21 G or 22 G needle, particularly when endobronchial ultrasound (EBUS)-TBNA and/or rapid-on-site evaluation (ROSE) are being used [24-27,35,49,50].

Thus, choosing needle size is at the discretion of the operator and depends on the goal of biopsy, pathology, cytopathology, or tissue culture. We typically use a 19 G needle for core biopsy material when large tissue samples are needed to evaluate tissue architecture (ie, suspected lymphoma); a 21 or 22 G needle is typically used when the goal of TBNA is cellular material for cytopathology and tissue culture or when safety is an issue (eg, biopsy of 4L station close to vascular structures). The size of needle typically used to perform TBNA in specific patient populations is shown in the table (table 4).

Technique — TBNA can be performed during conventional bronchoscopy (conventional-TBNA) or under real-time image-guidance using a bronchoscope with endobronchial ultrasound capability (EBUS-TBNA).

TBNA can be safely and successfully performed for endobronchial lesions encountered during routine flexible bronchoscopy and is recommended as the initial sampling modality given the lower complication rate and high diagnostic yield when compared to endobronchial brush and forceps biopsy.

During insertion of the catheter, the flexible bronchoscope should be kept as straight as possible, with its distal tip in the neutral position in order to prevent damage to the working channel by the needle. The beveled end of the needle should be kept within the metal hub of the catheter during its passage through the working channel [51,52].

The needle should be advanced and locked into place only when the metal hub is visible beyond the tip of the flexible bronchoscope. The catheter is then retracted, and the tip of the needle kept distal to that of the flexible bronchoscope. The bronchoscope is then advanced to the target area, and the tip of the needle is anchored in the intercartilaginous space so as to penetrate the airway wall in as perpendicular manner as possible, with a needle to wall angle >45 degrees (figure 3). Protruding the catheter more than a few millimeters beyond the tip of the bronchoscope may kink the catheter, leading to difficulty in attaining proper needle position, penetration, and aspiration. Using the bronchoscope to splint the distal end of the catheter further facilitates sampling.

The techniques that may be used to insert the needle through the airway wall are shown in the figure and described in the table (figure 4A-C and table 5). These techniques may be used alone or in combination for a successful penetration of the needle through the tracheobronchial wall. The needle should be inserted to its fullest extent with the metal hub flush against the tracheobronchial wall (figure 5).

With the needle inserted, suction is applied at the proximal side port using a 20 to 60 mL syringe. Aspiration of blood indicates inadvertent penetration of a major intrathoracic vessel; if this occurs, suction is released, the needle is retracted, and a new site is selected for aspiration. If no blood is aspirated, the catheter is agitated to and fro to obtain cytology specimens and is moved in and out by approximately 5 millimeters to obtain histology specimens, with continuous suction, in an attempt to shear off cells from the mass or the lymph node. Suction is released while the needle is within the node and the needle is then withdrawn from the target. The tip of the bronchoscope is straightened and the needle assembly removed from the bronchoscope in a single, smooth motion [52].

Use of the 19 G needle assembly and a variation of the technique used to obtain cytology specimens allows for obtaining histology samples via TBNA. The assembly is passed until only the distal metal hub is seen beyond the tip of the bronchoscope, while taking the same precautions as with the cytology needle. Once the metal hub is visible, the 19 G needle is advanced and locked into place beyond the metal hub (figure 6).

Under continuous suction, the 19 G needle is moved to and fro four to five times to obtain a core of tissue. The needle is flushed into the preservative to obtain the specimen. Successful sampling is indicated by visualizing a definite core of tissue in the preservative container.

Submucosal lesions may be sampled by incomplete penetration of the airway wall at an acute angle. Sampling of endobronchial lesions may be performed by fully embedding the needle in the lesion and agitating to and fro repeatedly, with continuous suction. Endobronchial needle aspiration is considered by many bronchoscopists to be the safest way to sample a visible lesion (when compared with forceps or brushing), and provides excellent cellular material and serves as a “litmus test” to assess the degree of bleeding one may encounter with forceps or brush. Sampling of peripheral lesions can be performed under fluoroscopic, electromagnetic, robotic, or radial-EBUS guidance, placing the metal hub of the needle against the lesion before pushing the needle into the lesion [52].

The number of aspirations performed at each target site impacts diagnostic yield. One study found that the diagnostic yield increased incrementally until the seventh aspiration [53]. A subsequent study reported similar results [54]. However, the latter study also indicated that three passes were sufficient only when alternate sampling sites were available and that four or five passes should be performed at nodal stations critical for staging. If ROSE is not available, guidelines suggest three to five passes for diagnosis, with an additional pass for ancillary studies (eg, molecular and immunohistochemical analysis) [55]. If ROSE is available, then clinical judgement can be used to determine the number of needle passes.

Tissue preparation — Proper handling of the obtained specimen is a critical and underappreciated aspect of the procedure. The specimen for cytology is prepared by using air from an empty syringe to spray the specimen onto the slide, smearing it using another slide, and immediately placing it in a 95 percent alcohol solution [56]. Alternatively, one of the slides can be prepared for rapid onsite cytologic evaluation (ROSE). Delay of even a few seconds may result in drying artifacts on the cells. The specimen is then flushed from the needle using 3 mL of saline or Hank's solution, and sent to the cytology lab for further processing.

Specimens are sent for cell block preparation, which provides either cytologic or histologic examination, depending upon the size and integrity of the sample [23]. When core specimens are desired, at least two satisfactory specimens should be obtained at each location; multiple passes may be required [57].

Tissue interpretation — ROSE of the specimen by a cytopathologist/cytotechnologist for sample adequacy has been demonstrated to increase diagnostic yield [27,33,34,58]. Some of the criteria that may be used to judge the adequacy of a sample obtained for cytology are listed in the table (table 6). Specimens containing rare malignant cells rather than clumps of malignant cells should not be classified as malignant and should not be relied upon for definitive diagnosis [59-62]. The steps that may be taken to reduce false positive results and avoid overstaging of lung cancer are highlighted in the table (table 7).

Trained assistant — A trained assistant schooled in the intricacies of TBNA is essential to ensure smooth performance of the procedure and proper tissue handling, which are critical to achieving a high yield [52].

COMPLICATIONS — Complications of transbronchial needle aspiration (TBNA) are procedure-related (eg, bleeding, pneumothorax) and sedation-related (eg, hypotension, respiratory failure), which are discussed separately. (See "Flexible bronchoscopy in adults: Preparation, procedural technique, and complications", section on 'Complications' and "Procedural sedation in adults in the emergency department: General considerations, preparation, monitoring, and mitigating complications", section on 'Anticipating and mitigating Complications'.)

Complications following TBNA are uncommon if appropriate precautions are taken and the proper technique is employed, and TBNA is considered one of the safest bronchoscopic sampling modalities. A coagulation profile is not needed prior to TBNA in the absence of a history of a bleeding diathesis [63]. Although a variety of complications related to TBNA have been reported (table 8), damage to the working channel of the bronchoscope is by far the most common. This complication is more common when a fixed or a 19 gauge (G) needle is used, and great care needs to be taken while manipulating the apparatus through the bronchoscope [51,64,65] (figure 7).

The incidence of fever and bacteremia has been debated, and no firm recommendations can be made regarding antibiotic prophylaxis [66-68] . Transient bacteremia six hours after the procedure with prompt defervescence after antibiotic therapy has been reported. After TBNA of a subcarinal mass, one patient experienced a purulent pericarditis with polymicrobial mouth flora and required pericardiocentesis and catheter drainage in addition to the antibiotics [69].

Marked bleeding has not been reported, even in patients receiving anticoagulation [70,71]. Oozing of a minimal amount of blood from the puncture site may be encountered; the source is usually a dilated blood vessel in the tracheobronchial wall rather than invasion of a major mediastinal vascular structure. Although patients with superior vena caval obstruction are believed by some to be at increased risk of bleeding, a retrospective study of 27 patients reported no major bleeding complications [72]. The superior vena cava may actually be displaced anteriorly by enlarged paratracheal nodes, making inadvertent puncture less likely.

Less frequent complications include pneumothorax, pneumomediastinum, hemomediastinum, mediastinitis, and needle fracture [6,40,73-76]. One patient with an elevated right hemidiaphragm underwent inadvertent liver biopsy but did not suffer any adverse clinical consequences [77].

MEDIASTINAL AND HILAR LYMPHADENOPATHY — Transbronchial needle aspiration (TBNA) has been most extensively employed in the diagnosis and staging of lung cancer, although the technique is also useful in a number of other settings, including the diagnosis of benign pathology (eg, sarcoidosis or infection). In most circumstances, TBNA in this setting is done under ultrasound guidance (ie, endobronchial ultrasound [EBUS-TBNA]).

Lung cancer — Accurate staging of lung cancer with preoperative detection of nodal metastases is critical to planning optimal treatment, including resection with curative intent and stereotactic body radiation therapy. It has been shown that "understaging" is associated with a worse survival [78]. (See "Overview of the initial evaluation, diagnosis, and staging of patients with suspected lung cancer" and "Tumor, node, metastasis (TNM) staging system for lung cancer".)

Lymph node enlargement on computed tomography (CT) scan or uptake of fluorodeoxyglucose on positron emission tomography (PET) does not constitute proof of malignant spread. Likewise, a negative PET scan cannot rule out nodal disease. The major limitation of imaging is that neither CT nor PET can diagnose or stage non-small cell lung cancer (NSCLC) with a high degree of accuracy. A 2013 meta-analysis reported a median sensitivity for CT and PET scan of 55 and 77 percent, respectively [79]. Thus, CT and PET scanning are most useful for directing the biopsy of suspicious nodes but do not provide definitive staging information [43,80]. (See "Overview of the initial evaluation, diagnosis, and staging of patients with suspected lung cancer" and "Tumor, node, metastasis (TNM) staging system for lung cancer".)

The options available for lymph node sampling include the following:

●Minimally-invasive modalities (see "Procedures for tissue biopsy in patients with suspected non-small cell lung cancer", section on 'Endoscopic and percutaneous procedures')

●Surgical techniques (see "Procedures for tissue biopsy in patients with suspected non-small cell lung cancer", section on 'Surgical staging procedures')

The minimally-invasive modalities are conventional TBNA, and EBUS-TBNA to obtain tissue to stage the hilum/mediastinum in patients with lung cancer. Compared with surgical techniques (eg, cervical mediastinoscopy, video-assisted thoracoscopy), the advantages of c-TBNA or EBUS-TBNA are the following [16,43,70]:

●TBNA, in particular EBUS-TBNA, can access more lymph node stations (2, 3, 4, anterior and posterior 7, 10, 11) than cervical mediastinoscopy (2, 3, 4, and anterior 7) (figure 1) [81-87].

●TBNA can access both sides of the tracheobronchial tree (right and left), compared with video-assisted thoracoscopy which can only access one side of the mediastinum.

●TBNA, as an adjunct to bronchoscopy, can be performed under moderate sedation.

●TBNA can be performed in situations where mediastinoscopy is difficult or impossible, such as in patients with tracheostomy, those with severe cervical deformities, and in those who have either had an initial staging mediastinoscopy or received prior chest radiation therapy.

●TBNA, as an adjunct to bronchoscopy, provides tissue for diagnosis and staging while simultaneously allowing evaluation of the endobronchial tree.

In patients with lymphadenopathy from suspected lung cancer, factors that increase the sensitivity of c-TBNA include:

●Airway involvement (visible tumor or erythema) [16,43,44,88,89]

●Subcarinal lymph node size greater than 2 cm [16,89]

●Suspected small cell carcinoma [14,41,88,90,91]

EBUS-TBNA is superior to conventional bronchoscopic-TBNA and has become the preferred first-step procedure for staging hilar/mediastinal lymph nodes in patients with suspected lung cancer [79,81-83,92]. The accuracy of EBUS-TBNA has been shown to be as high as 96 percent, and again, can sample both mediastinal and hilar stations (2R, 2L, 3p, 4R, 4L, 7, 10R, 10L, 11R, 11L) (figure 1) [81-87,93]. High operator proficiency and multiple passes with rapid on-site cytologic evaluation (ROSE) may enhance its diagnostic accuracy [24,25,27-30]. However, it cannot be used for nodes that are not within its reach (stations 5, 6, 8, 9, and the intrapulmonary station 12 nodes).

It is crucial that the highest lymph node stations are sampled first in order to prevent inadvertent up-staging if the same needle is used for subsequent lymph node stations. The mantra of the bronchoscopist should be “N3, then N2, then N1.” It has been shown that experienced bronchoscopists do a better job with lung cancer staging, especially following the “N3, N2, N1” rule [94]. Additionally, as it has been shown that patients with multi-station N2 disease have a worse survival than patients with single-station N2 disease, samples should be obtained from all nodes >5 mm in diameter [95,96].

It should be noted that the false negative rates associated with c-TBNA necessitate additional tissue sampling (ie, surgical sampling) when testing is negative for cancer. This may not be required if the nodes are negative following EBUS-TBNA, however, this decision should be made based on an assessment of local false-negative rates and in conjunction with thoracic surgery. (See "Procedures for tissue biopsy in patients with suspected non-small cell lung cancer", section on 'Diagnostic and staging accuracy' and "Selection of modality for diagnosis and staging of patients with suspected non-small cell lung cancer", section on 'Approach to the patient'.)

Lymphoma — The clinical utility of TBNA in the diagnosis of lymphoma has been somewhat limited, since this may require larger samples of nodal tissue than are normally obtained by a 19 or 22 gauge needle. However, the diagnosis of lymphoma, using both cytology and histology needles, has been reported, although the yield of the procedure in this setting cannot be stated with certainty [46,62,64,71,97-102]. The availability of the 19 GA histology needle, along with the use of flow cytometry to enhance diagnostic yield, may change this assumption.

Several observational series report the sensitivity and specificity of EBUS-TBNA in patients with suspected new or recurrent lymphoma [98-103].

●One meta-analysis of 14 observational studies reported an overall sensitivity of 66 percent and specificity of 99 percent [103]. A similar sensitivity and specificity was reported for patients with a new diagnosis of lymphoma, while for patients with recurrence the sensitivity was slightly better at 78 percent. Use of rapid on-site evaluation and flow cytometry in those with recurrence may have led to the improved sensitivity.

●In a prospective study, 98 patients with isolated mediastinal lymphadenopathy underwent EBUS-TBNA [100]. EBUS-TBNA definitively diagnosed lymphoma with a sensitivity and specificity of 57 and 100 percent, respectively. Sixty-six percent of the patients were diagnosed with lymphoma by EBUS-TBNA, but 25 percent required surgical biopsy to determine the specific type due to the small amount of tissue obtained by EBUS-TBNA.

●The utility of EBUS-TBNA was investigated prospectively in 100 patients in whom a new diagnosis or recurrent mediastinal lymphoma was suspected [98]. EBUS-TBNA correctly diagnosed lymphoma in 60 percent of patients with a sensitivity and negative predictive value of 89 and 83 percent, respectively. In the population of patients with suspected relapse, the diagnostic yield for lymphoma was higher than in those with a newly-suspected condition (100 versus 88 percent). The sample was considered adequate for both grading (high-grade non-Hodgkin, low-grade non-Hodgkin, and Hodgkin lymphoma) and management in 84 percent of cases.

The problem of lower diagnostic accuracy may be overcome by rapid on-site evaluation, since EBUS-TBNA provides sufficient sample for definitive primary diagnosis and classification of malignant lymphoma using immunohistochemistry, immunophenotyping, and/or fluorescence in situ hybridization (FISH). Rapid on-site specimen assessment is invaluable for appropriate assignment of the sample to ancillary studies [101].

Sarcoidosis — In the past, patients suspected to have sarcoidosis with negative transbronchial lung biopsy were referred for mediastinoscopy. The availability of TBNA offers a less invasive, safer, and more economical alternative for obtaining a pathologic diagnosis from hilar/mediastinal lymph nodes. Perhaps most importantly, granulomas seen on ROSE can obviate the need for parenchymal lung biopsy, decreasing the risk of bleeding and pneumothorax.

Several reports have confirmed the diagnostic value of TBNA for the biopsy of mediastinal lymph nodes in patients with suspected sarcoidosis [104-114]. As an example, one study of 258 patients with suspected sarcoidosis found that the diagnostic yield increased from 66 to 78 percent when TBNA was added to transbronchial forceps biopsy [107]. A second randomized multicenter study (GRANULOMA) of 304 patients with suspected stage I/II sarcoidosis reported that compared to transbronchial biopsy, TBNA of mediastinal nodes by endobronchial or esophageal ultrasound resulted in a higher diagnostic yield (80 versus 53 percent) [110]. This study should be interpreted with caution given that not all patients with stage I sarcoidosis need histologic diagnosis and not all bronchoscopists are proficient in both procedures (bronchoscopy and esophageal endoscopy).

Procedures used to establish the diagnosis of sarcoidosis are discussed separately. (See "Clinical manifestations and diagnosis of sarcoidosis", section on 'Endoscopic ultrasound-guided needle aspiration or cryobiopsy'.)

Infection — A number of anecdotal reports of infection diagnosed via TBNA have appeared. TBNA has been used to diagnose histoplasmosis, Pneumocystis jirovecii pneumonia, and cryptococcal infection in patients with AIDS [47]. In addition, the diagnosis of mediastinal mycobacterial adenitis (due to either Mycobacterium tuberculosis or M. avium-intracellulare) has been described in immunocompetent, as well as immunocompromised patients [47,115-124].

A 1998 report highlights the usefulness of this technique in patients with HIV [125]. Forty-one HIV-positive patients with intrathoracic adenopathy of unclear etiology underwent 44 cTBNA (19 GA) procedures, and 23 patients were demonstrated to have mycobacterial disease; aspirates showed smears positive for acid fast bacilli (AFB) in 11 (48 percent), 14 (61 percent) specimens grew mycobacteria, and caseous necrosis or necrotizing granulomatous lesions were seen in 15 (65 percent). In 11 of 23 procedures (48 percent), TBNA was the exclusive means of diagnosing mycobacterial disease. No major complications were reported. The report found TBNA to be less successful in patients with lymphoma and Kaposi's sarcoma.

In another retrospective analysis of 249 patients who underwent EBUS-TBNA, heterogeneous echotexture (53 versus 13 percent) and coagulation necrosis (26 versus 3 percent) were more commonly found in lymph nodes infected with tuberculosis with a positive tuberculin test; EBUS-TBNA had a specificity of 98 percent and a positive predictive value of 91 percent for tuberculosis [124].

Other — Submucosal needle aspiration proximal to an endobronchial tumor to detect local spread may help predict the line of surgical resection in patients with non-small cell carcinoma [126]. Transbronchial needle aspiration has identified leiomyoma of the esophagus [47], metastatic uterine rhabdomyosarcoma [127], sclerosing hemangioma [128], and malignant pleural mesothelioma [129], while endobronchial aspirates have diagnosed both carcinoid tumors [130,131] and malignant melanoma [132].

Mediastinal masses have been diagnosed and therapeutically aspirated using the transbronchial route [133-137]. One report described a patient with a right paratracheal mass on CT scan that was suspicious for malignancy; TBNA revealed serosanguineous fluid suggestive of a sterile abscess, and there was no recurrence on later scans following the aspiration [138]. Decompression of a subcarinal cyst using TBNA permitted safe anesthesia and the subsequent resection of the cyst in another report [139]. It should be noted that by passing the needle through the airway wall, an otherwise sterile bronchogenic cyst can become infected.

AIRWAY LESIONS — Due to the high yield (67 to 100 percent) of forceps biopsy for diagnosing visible endobronchial lesions suspicious for lung cancer, the indications of transbronchial needle aspiration (TBNA) are unclear [91,140-145]. However, TBNA has sometimes yielded the only diagnostic sample which may be due to the ability of TBNA to not only access tumor that is visible (especially those of the right upper lobe) but also tumor that is peribronchial, as well as submucosal in nature [43,88,89,91,146]. Overall, reports suggest that for central lesions in patients with suspected cancer, TBNA has a diagnostic yield between 70 and 96 percent [9,10,14,140-145,147].

Clinical situations in which TBNA of an endobronchial mass can be useful include:

●A visible lesion with surface necrosis which could lead to a negative forceps biopsy result (TBNA can help establish a diagnosis by obtaining samples from within the viable tumor mass)

●An endobronchial lesion which is likely to bleed, such as a carcinoid tumor or renal cell carcinoma [130,131]

●Suspected small cell carcinoma (extensive crush artifact on forceps biopsy specimens can decrease the diagnostic yield) [148]

Peribronchial and submucosal disease — The yield of conventional procedures such as forceps biopsy and brushing tends to be much lower in submucosal and peribronchial disease than for exophytic lesions because:

●The abnormal lesion may be covered by normal epithelium, causing suboptimal sampling

●Submucosal infiltration by tumor may make tissues firmer, causing the forceps to slide off the lesion

●Peribronchial lesions are inaccessible to the biopsy forceps by virtue of being located outside the airway

Under such circumstances, the addition of TBNA could increase the diagnostic yield [9,10,14,149]. One study involving 31 patients with submucosal and peribronchial disease found that the sensitivity of biopsies obtained by forceps, TBNA, a combination of both forceps and TBNA, and a combination of forceps biopsy, bronchial brushing and washing, and TBNA was 55 percent, 71 percent, 89 percent, and 97 percent, respectively [9].

LUNG NODULES OR MASSES — Transbronchial needle aspiration (TBNA) of parenchymal lung nodules and masses has emerged as an extremely useful diagnostic technique.

Solitary pulmonary nodules (SPN) can be classified into four types according to the tumor-bronchus relationship (figure 8) [150]:

●The bronchial lumen is patent up to the tumor

●The bronchus is contained in the tumor mass

●The bronchus is compressed and narrowed by the tumor but the bronchial mucosa is intact

●The proximal tree is narrowed by peribronchial or submucosal spread of the tumor or by an enlarged lymph node

TBNA is particularly useful for increasing the diagnostic yield of flexible bronchoscopy in lesions in which the bronchus is compressed and narrowed by the tumor but the bronchial mucosa is intact, or in which the proximal bronchial tree is narrowed by peribronchial or submucosal spread of the tumor or by enlarged lymph nodes. In such situations, the submucosal or peribronchial location makes the tumor relatively inaccessible to other procedures such as washing, brushing, or transbronchial biopsy. Details regarding the diagnostic evaluation of SPNs are discussed separately. (See "Diagnostic evaluation of the incidental pulmonary nodule".)

PERIPHERAL NODULES OR MASSES — Traditionally, smaller peripheral lesions are not easily accessed by conventional bronchoscopy that relies on conventional transbronchial biopsy to access suspicious lesions. However, a 2013 meta-analysis indicates that the yield from bronchoscopic TBNA is higher (median sensitivity 63 percent) in peripheral lesions than from either forceps biopsy alone or a combination of conventional biopsy, brushing, and lavage procedures [151]. Lesion characteristics predicting the best diagnostic yields include:

●Size greater than 2 cm (80 versus 33 to 58 percent)

●Concurrent mediastinal disease (89 versus 46 percent)

●The lesion being a hematogenous metastasis

The yield from a combination of TBNA plus forceps biopsy may be as high as 75 percent, although the yield is lower than with transthoracic needle aspiration (TTNA) [73].

Combining TBNA with image-guidance may increase the diagnostic yield of TBNA for peripheral lung nodules:

●One systematic review of 18 studies reported a higher diagnostic yield with fluoroscopic-guided TBNA when compared with blind transbronchial biopsy (60 versus 45 percent) [152]. The presence of the computed tomography bronchus sign, large malignant lesions (>3 cm), and the employment of rapid on-site evaluation (ROSE) predicted a higher yield with fluoroscopy guidance.

●In one single series of 467 nodules, using radial probe EBUS, a radial probe position within the target lesion enhanced the diagnostic yield compared with a radial probe position that was adjacent to the lesion (84 versus 48 percent) [153].

●In contrast, computed tomography (CT) fluoroscopy has not been shown to increase the diagnostic yield of TBNA for peripheral lung nodules. A trial that randomly assigned 50 patients with suspected lung cancer to undergo CT fluoroscopy-guided TBNA or conventional TBNA found no difference in the sensitivity of detecting malignancy in peripheral lesions (71 versus 76 percent for conventional TBNA) [154].

Based upon this data, in the hands of an experienced operator, we typically consider radial probe endobronchial ultrasound (RP-EBUS) and TBNA as a reasonable first-choice biopsy procedure for peripheral nodules and masses ≥3 cm. The use of radial probe EBUS-TBNA for lesions <3 cm is less sensitive but can be used at the discretion of the physician. The biopsy of solitary pulmonary nodules and use of image-guided bronchoscopy techniques for the biopsy of peripheral nodules are discussed separately. (See "Diagnostic evaluation of the incidental pulmonary nodule", section on 'Nonsurgical biopsy' and "Image-guided bronchoscopy for biopsy of peripheral pulmonary lesions".)

SUMMARY AND RECOMMENDATIONS

●The most common application of transbronchial needle aspiration (TBNA) is the diagnosis and staging of lung cancer. However, TBNA can also be used to sample mediastinal, hilar, endobronchial, submucosal, peribronchial (when extrinsic compression exists), or peripheral lesions (table 1). (See 'Indications' above.)

●Computed tomography (CT) is routinely performed prior to the bronchoscopy and TBNA in order to evaluate the relationship of the tracheobronchial tree to the surrounding structures, including lymph nodes and blood vessels. This maximizes diagnostic yield and safety. (See 'Computed tomography' above.)

●Smaller needles (eg, 20 to 22 gauge) are used to obtain cytology specimens, whereas larger needles (eg, 19 GA) are necessary to obtain a core of tissue for histology. (See 'Type of needle' above.)

●Proper handling of the specimen is critical. The first specimen is sprayed from the needle onto a glass slide using air, smeared using another slide, and then placed in 95 percent alcohol solution. Delays can result in drying artifacts. Subsequent specimens are flushed from the needle using saline or Hank's solution. On-site evaluation by a cytopathologist may improve the diagnostic yield. (See 'Tissue preparation' above.)

●A coagulation profile is not needed prior to TBNA in the absence of a history of bleeding diathesis. (See 'Complications' above.)

●To perform TBNA, a needle-containing catheter is advanced through the flexible bronchoscope until the metal hub of the catheter is visible. The needle is then advanced from the tip of the catheter and locked into position. Next, the catheter is retracted until only the needle tip is visible. The bronchoscope is then advanced until the tip of the needle is anchored in its target, suction is applied, and the catheter is agitated (if no blood was aspirated). The needle-containing catheter is removed after suction is released. (See 'Technique' above.)

●Endobronchial ultrasound (EBUS)-TBNA is superior to conventional bronchoscopic-TBNA and has become the preferred first-step procedure for staging mediastinal and hilar lymph nodes in patients with suspected lung cancer. Its diagnostic sensitivity may be enhanced when multiple passes and rapid onsite evaluation (ROSE) is concurrently used. (See 'Mediastinal and hilar lymphadenopathy' above.)

●For patients with airway lesions, TBNA has a high diagnostic yield. TBNA has also been shown to improve sampling of peripheral nodules and masses. (See 'Airway lesions' above and 'Lung nodules or masses' above.)