Group
Description
Definition (by chest CT scan)
A
Mediastinal infiltration
Tumor mass within the mediastinum such that discrete lymph nodes cannot be distinguished or measured
B
Enlarged discrete mediastinal nodes
Discrete mediastinal nodes ≥1 cm in short-axis diameter on a transverse CT image
C
Clinical stage II or central stage I tumor
Normal mediastinal nodes (<1 cm) but enlarged N1 nodes (>1 cm) or a central tumor (within proximal one-third of the hemithorax)
D
Peripheral clinical stage I tumor
Normal mediastinal and N1 nodes (<1 cm) and a peripheral tumor (within outer two-thirds of hemithorax)
Fig. 23.1
The American College of Chest Physicians intrathoracic radiographic (CT scan) categories for lung cancer. (a) Mediastinal infiltration by tumor. (b) Enlarged discrete N2/N3 nodes. (c) A central tumor or a tumor with enlarged N1 nodes but a normal mediastinum. (d) A peripheral small tumor with normal-sized lymph nodes
Role of Positron Emission Tomography (PET) Imaging
Whole-body PET imaging in preoperative staging of lung cancer has been shown to increase identification of patients with mediastinal and extrathoracic disease compared to conventional staging [10]. In addition, PET improves discrimination between N0/N1 and N2/N3 diseases [11]. However, one of the downsides to increasing sensitivity in detecting occult metastases is incorrectly upstaging the disease in more patients, potentially withholding possible curative management. In five randomized controlled trials (RCTs) involving PET scans, this would have occurred in 5–42%; however, in these studies, the requirement of confirmation of these abnormal findings prevented this [6].
There have been several small randomized controlled trials with conflicting results. Two RCTs utilizing PET scanning to accurately stage lung cancer found a significant reduction, from approximately 40 to 20%, in the number of futile thoracotomies performed (defined as thoracotomy with the finding of a benign lesion, pathologically confirmed lymph node involvement, unresectable disease, discovering metastatic disease, or recurrence or death from any cause within 1 year) but did not affect overall mortality [12, 13]. However, this was not replicated in an RCT evaluating patients with stage I–II NSCLC [14].
PET scanning is not a conclusive test, and suspicious lesions do not obviate the need for histologic evaluation of the mediastinal lymph nodes for accurate staging. In an analysis of a previously reported trial, PET-CT had a 70% sensitivity and 94% specificity; however, of the 22 patients with PET-positive mediastinal nodes, 8 did not have tumor. The positive predictive value and negative predictive value are 64% and 95%, respectively, corroborating the need for pathologic confirmation with a positive PET-CT [15]. The frequency of false-negative mediastinum on PET-CT has been reported to range between about 15 and 28% [11, 16]. Furthermore, the PET imaging results should be interpreted in relation to lymph node size. The negative predictive value in normal-sized lymph nodes (<10 mm) is significantly higher (96% vs. 70%), and the positive predictive value is lower (43% vs. 71%). Thus, in patients without mediastinal lymphadenopathy, a negative PET-CT is highly valid and patients may proceed to surgery unless they have a central tumor. However, the false-negative rate is considerable in enlarged lymph nodes without FDG uptake (30%) [11]. In such cases, needle techniques to assess the mediastinum (EUS-FNA, EBUS) would be the most rational next step.
Invasive Approach to Staging the Mediastinum
The initial radiographic staging helps the clinician categorize the patient into one of the four radiographic groups previously discussed (Table 23.1). This classification helps guide the selection of invasive test and defines the performance characteristics of these tests. The former relies on anatomic factors (i.e., location, accessibility, and size of the nodes), while the latter is dependent on the local availability and operator experience.
Transthoracic Needle Aspiration
Leyden performed the first transthoracic needle aspiration (TTNA) to confirm a pulmonary infection in 1883 [17]. Since then, TTNA has become an efficacious procedure with no significant morbidity mainly due to two major factors, advances in imaging technology and improvements in histopathology [18].
It involves passing a needle percutaneously under image guidance to either aspirate or biopsy (TTNB) tissue. The transition from fluoroscopy to computed tomography led to improved visualization of even smaller lesions that can be approached with a greater margin of safety and accuracy.
The sensitivity of TTNA for staging the mediastinum has been reported to be high at 91% in a meta-analysis of five studies involving 215 patients [19]. These studies may be inherently biased because the study patients enrolled had bulky mediastinal disease and extrapolation of these results to patients with lesser amounts of mediastinal spread for staging purposes may be inappropriate. In addition, due to the proximity of the major thoracic vessels and the heart to the mediastinal lymph nodes, TTNA is mostly limited to the superior mediastinal lymph nodes. Iatrogenic pneumothorax is the most common complication, averaging 10% when using a “protective technique” but may go up to 60% if the lung parenchyma is traversed [20]. These factors limit the use of TTNA in staging the mediastinum especially in COPD patients with severe emphysema.
Transbronchial Needle Aspiration
Transbronchial needle aspiration (TBNA) was first reported by Dr. Schieppati in 1949 although it was in 1978 when Wang and colleagues described the diagnosis of a paratracheal mass by TBNA through a rigid bronchoscope [21].
TBNA for mediastinal staging is performed through the bronchoscope. It involves passing the needle catheter, which comes in different sizes, through the working channel of the bronchoscope and then directed to the target lesion. The needle is then passed through the bronchial wall, and material is aspirated for tissue sampling. It can be performed as an unguided procedure during bronchoscopy or under image guidance using a bronchoscope with endobronchial ultrasound or electromagnetic navigational capability. It is used most commonly to assess subcarinal lymph nodes and less frequently with paratracheal lymph nodes due to difficulty in adequately directing the bronchoscope and needle. Rapid on-site cytological evaluation of the tissue obtained is a cost-effective method to improve the yield, eliminating unnecessary passes during a procedure [22].
The overall median sensitivity was moderate at 78% (range 14–100%), and the negative predictive value was 77% in systematic review evaluating 2408 patients. The reported specificity and false-positive rates were 100% and 0%, respectively [6]. The patients included in the studies mainly had N2/N3. As such, these results can be reliably applied to patients with bulky mediastinal disease; however, the high false-negative rate makes TBNA less useful for staging the mediastinum in patients with normal-sized lymph nodes. A negative TBNA therefore cannot effectively rule out mediastinal nodal involvement, and additional staging procedures should be performed. In a comparative study directly evaluating the accuracy of TBNA against endobronchial ultrasound fine needle aspiration (EBUS-FNA) and endoscopic ultrasound fine needle aspiration (EUS-FNA), TBNA was less sensitive when individually compared to EBUS-FNA and EUS-FNA in identifying mediastinal involvement (36% vs. 69%). This effect was seen across different subgroups including individuals with stations favorable to TBNA, such as a PET-positive subcarinal node or enlarged subcarinal node [23].
Endoscopic Ultrasound with Needle Aspiration
Endoscopic ultrasound with fine needle aspiration (EUS-FNA) for cytologic diagnosis was initially used in pancreatic diseases but was first described for mediastinal node aspiration in 1993 [24]. It is generally safe and well tolerated under local anesthesia and conscious sedation.
EUS is performed using an endoscope with an ultrasound transducer at the tip, also known as an echoendoscope . There are two types of echoendoscopes, radial and curvilinear; the former provides a 360° view, while the latter is used for aspiration and provides a 180° view of the structures adjacent to the gastrointestinal tract. The inferior and posterior mediastinum is localized through the esophageal wall providing direct visualization of stations 4L, 5, 7, 8, and 9. However, EUS cannot evaluate the anterior mediastinum and right paratracheal lymph node stations due to air interference from the trachea; thus, complete visualization of the mediastinum cannot be achieved with EUS alone. The left adrenal gland can also be sampled when it is enlarged and PET avid, potentially offering diagnosis and complete staging with a reported sensitivity and NPV of 86% and 70%, respectively [25]. On the other hand, the diagnostic yield of EUS-FNA for detecting malignant disease in liver lesions is significantly lower at 58% [26].
In a meta-analysis looking at 18 eligible studies, the utility of EUS-FNA for staging mediastinal lymph nodes (N2/N3) was evaluated [27]. The pooled sensitivity was 83%, and the pooled specificity was 97%. The sensitivity and specificity were slightly higher at 90% and 97%, respectively, among patients with mediastinal lymphadenopathy on imaging. Comparatively, among patients enrolled without enlarged lymph nodes, the pooled sensitivity was significantly lower at 58% [27]. Furthermore, in a systematic analysis of 2433 patients, the median sensitivity and specificity were comparable at 89 and 100%; however, most of these patients were generally selected because they had nodal disease amenable to EUS-FNA (6). Due to its modest negative predictive value at 78%, a nondiagnostic result will need to be further evaluated [27].
In a small RCT evaluating the value of EUS-FNA in mediastinal staging, EUS-FNA together with rapid on-site evaluation by a cytopathologist was found to reduce the need for surgical staging by 68% and was found to have a higher but not statistically significant sensitivity (93% vs. 73%) in detecting malignant disease. The complication rate was less in the EUS group at 0% vs. 5%, although this was not statistically significant [28].
Endobronchial Ultrasound with Needle Aspiration
Endobronchial ultrasound (EBUS) is performed during bronchoscopy that uses ultrasound to visualize surrounding structures within the airway wall and mediastinum. There are two types of EBUS, the radial probe EBUS (RP-EBUS) and the convex probe (CP-EBUS). Briefly, RP-EBUS has a higher resolution such that airway structure and parenchymal lesions are visualized in better detail for subsequent TBNA; however, it cannot be used to biopsy targets in real time. The capacity to allow for real-time ultrasound-guided TBNA was first reported in 2004, with the incorporation of a convex ultrasound probe [29].
The RP-EBUS is performed by placing a conventional bronchoscope in the area of interest and by inserting a radial ultrasound probe through the working channel followed by inflating the probe balloon with water and subsequent TBNA. The same technique is used for CP-EBUS, providing a view that is 35° forward oblique. Color flow and Doppler features allow identification of vascular and cystic structures.
EBUS can access a wide range of mediastinal and hilar lymph nodes including 2R, 2L, 3P, 4R, 4L, 7, 10R, 10L, 11R, and 11L (Fig. 23.2).
Fig. 23.2
Regional lymph node stations for lung cancer staging. Printed with permission from Naruke et al. and the ATS/North American LCSG
The overall median sensitivity of EBUS-NA in mediastinal staging is reported to be at 89% in a systematic review of 2756 patients, with values ranging from 46 to 97%. The median NPV was 91% [6]. Most of the studies in this review included patients with bulky lymphadenopathy, mostly radiographic group B and some A and C. However, two studies evaluated the performance of EBUS-FNA in patients with normal mediastinum by CT scan and PET-CT, respectively. The prevalence of mediastinal disease was lower in the negative PET-CT group, likely due to the higher sensitivity of PET-CT to detect disease. Despite this, the negative predictive value was comparable in both groups at around 96% [30, 31].
Combined EUS-FNA and EBUS-FNA
EUS-FNA and EBUS-FNA have a complimentary diagnostic yield and can potentially allow complete access to all nodal stations with EUS providing access to posterior and inferior mediastinum as well as the evaluation of the left adrenal gland in selected cases and EBUS providing access to the anterior and superior mediastinal lymph nodes. In a systematic review of seven studies including 811 patients, the pooled median sensitivity and specificity were 91% and 100%, respectively [6]. Although the concept of performing both procedures in one setting is ideal, it is often difficult to achieve due to the uncoordinated use of EUS and EBUS in most centers.
Surgical Staging
The most common surgical modalities used for mediastinal staging of non-small cell lung cancer include standard cervical mediastinoscopy, video-assisted thoracoscopic surgery (VATS), and anterior mediastinotomy (Chamberlain procedure). Technique selection usually relies on clinical judgment, knowledge of diagnostic accuracy, operator proficiency, and local expertise.
Standard Cervical Mediastinoscopy (SCM)
Standard cervical mediastinoscopy is an invasive modality that is used to evaluate the superior and middle mediastinum for lung cancer staging. It was first described by Harken in 1954 when they inserted a laryngoscope into the mediastinum through a supraclavicular incision, and lymph node biopsies were taken [32]. Carlens from Sweden developed the pretracheal, suprasternal approach, as it is practiced today [33].
Its major advantage over minimally invasive techniques is direct visualization of nodes for sampling and dissection, thereby allowing relatively large biopsies to be obtained, yielding more than adequate samples for culture, immunohistochemical, and molecular analysis.
Mediastinoscopy is usually an outpatient procedure performed in the operating room under general anesthesia requiring around 1–2 days for recovery. A single institutional review of 2145 patients undergoing mediastinoscopy over 9 years showed an overall reported morbidity and mortality of 1.07 and 0.05%. The most frequent reported complications included recurrent laryngeal nerve injury, hemorrhage, tracheal injury, and pneumothorax [34].
A midline transverse incision is made immediately above the sternal notch providing access to pretracheal (1, 3), paratracheal (2R, 2L, 4R, 4L), anterior subcarinal (7), and occasional hilar (10R, 10L) stations (Fig. 23.2). Nodal stations that cannot undergo sampling using this technique include the posterior subcarinal (7), the inferior mediastinal (8, 9), the aortopulmonary window (5), the anterior mediastinal, and the lobar/interlobar (11–14) stations. The advent of video mediastinoscopy has permitted the concurrent use of multiple surgical instruments via the mediastinoscope which allows better visualization, more extensive sampling (including posterior station 7), and true mediastinal lymphadenectomy.
In a review of 9257 pooled patients, the overall median sensitivity of standard cervical mediastinoscopy compared to videomediastinoscopy and mediastinal lymphadenectomy was reported to be 78%, 89%, and 94%, respectively. The median NPV was 91%. The FN cases were predominantly nodal stations that were not accessible by the traditional mediastinoscopy and most likely affected operator diligence in node dissection and sampling. Ideally, the paratracheal and anterior subcarinal lymph nodes should be dissected routinely with at least one node sampled from each station especially in patients without clinical suspicion of node involvement [6].
Anterior Mediastinoscopy
Due to lymphatic drainage patterns, malignancies involving the left upper lobe have increased tendencies to involve nodes in the aortopulmonary window (station 5) and prevascular (station 6) stations (Fig. 23.2). Traditionally, these lymph nodes are classified as N2 nodes and are difficult to gain access to and cannot be reached by minimally invasive techniques or by standard cervical mediastinoscopy. In 1966, McNeill and Chamberlain described a technique of left anterior mediastinotomy via an incision in the second or third intercostal space just to the left of the sternum in order to gain access to these stations [35]. It has the advantage of providing access to left upper lobe tumors for simultaneous resection in the instance that there is no evidence of nodal involvement or distant metastasis.
In a systematic review, the reported median sensitivity of the Chamberlain procedure in the detection of metastatic nodal disease was approximately 71% among a pool of 238 patients while the median NPV was 91% [6]. There is a paucity of data comparing the efficacy of the Chamberlain procedure with VATS and other minimally invasive techniques, but in a small study of 112 patients, VATS was able to correctly identify malignancy in 100% patients compared to 83% patients that underwent Chamberlain procedure [36].
Extended Cervical Mediastinoscopy
Extended cervical mediastinoscopy (ECM) offers increased access to the sampling region usually offered by standard cervical mediastinoscopy whereby the subaortic and para-aortic lymph nodes can be reached in addition to nodal stations 1, 2, 3, 4, 7, and 10 (Fig. 23.2). The main advantage of ECM is its ability to provide additional access to the APW when staging tumors of the left upper lobe without the surgical risk associated with a Chamberlain procedure. However, ECM is not widely used and is only limited to a few centers that are specialized to this staging technique.
Video-Assisted Thoracic Surgery
The practical use of thoracoscopy was first reported by Dr. Jacobaeus in 1910 when he utilized a cystoscope for a thoracoscopic diagnosis of tubercular intrathoracic adhesions [37]. The traditional thoracoscope had several limitations including a limited view of the field that is also mostly restricted to the operator [38]. With the introduction of video-assisted imaging, the functional capacity of thoracoscopy has been significantly amplified. It magnifies the image with the aid of better instruments as well as shares the images with other clinicians performing the procedure.
It is a major surgical procedure that requires general anesthesia and hospital stay, incurring higher costs for training and equipment. It is typically utilized when other alternative modalities cannot access the tumor or are nondiagnostic. The major advantage of VATS is direct visualization of the lung and mediastinal structure including almost all nodal stations. It can potentially evaluate the extent of involvement by the primary tumor (T), mediastinal lymph node involvement (N), and pleural involvement (M1a) possibly offering intraoperative diagnosis, staging, and therapy. The major disadvantage of the latter is the increased morbidity and mortality associated with surgery.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
