| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Department of Pulmonary, Critical Care and Sleep Medicine, Medical University of South Carolina, Charleston, South Carolina
b Department of Thoracic Surgery, Medical University of South Carolina, Charleston, South Carolina
c Department of Gastroenterology and Hepatology, Medical University of South Carolina, Charleston, South Carolina
Accepted for publication July 9, 2007.
* Address correspondence to Dr Silvestri, Medical University of South Carolina, 96 Jonathan Lucas St, Ste 812 CSB, Charleston, SC 29425 (Email: silvestr{at}musc.edu).
| Drs Hoffman and Silvestri disclose that they have a financial relationship with Olympus America, Inc.
|
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
Methods: This was a retrospective analysis of 152 consecutive patients who underwent EBUS-TBNA with undiagnosed intrathoracic adenopathy or cancer staging as the primary indications. The procedures occurred between January 2005 and June 2006 at a single academic medical center. Of the 152 patients, 117 were included in the final statistical analysis after excluding those with benign disease diagnosed by EBUS-TBNA. Rapid on-site cytopathologic examination was used in all cases.
Results: Malignancy was identified in 113 patients, of which 67 (59.3%) had non-small cell lung carcinoma, and 20 (17.7%) underwent surgical resection. Four patients had benign diagnoses at surgical pathology. Only 1 surgical patient was found to have nodal metastasis at a lymph node station previously biopsied by EBUS-TBNA (negative predictive value, 97%). Compared with radiologic staging, EBUS-TBNA down-staged 18 of 113 (15.9%) and up-staged 11 (9.7%). Sensitivity was 98.7%, with 100% specificity. No major complications were associated with the procedure.
Conclusions: EBUS-TBNA is useful in accessing mediastinal and hilar lymph nodes for the diagnosis and staging of non-small cell lung cancer and other disorders of the mediastinum. Thoracic surgeons and pulmonologists are well positioned to use this tool in everyday practice.
Accurate staging of non-small cell lung cancer (NCLC) is a critical step that determines both the treatment modality and the prognosis. This is best accomplished with a multidisciplinary approach that involves surgical, pulmonary, oncologic, and radiologic input to establish whether curative surgical resection is possible.
Preoperative mediastinal lymph node staging separates initial operative versus nonoperative status. Patients without mediastinal nodal metastases are considered operative candidates, but patients with ipsilateral mediastinal or higher stage nodal metastases are treated primarily with chemotherapy and external beam radiation, although some patients receive preoperative neoadjuvant chemotherapy, followed by surgery.
Several invasive and noninvasive methods exist to stage intrathoracic lymph nodes. Noninvasive methods include computed tomography (CT) with intravenous contrast administration, 18-F-deoxyglucose positron emission tomography (FDG-PET), and simultaneous acquisition CT-PET fusion imaging. Given a false-positive rate as high as 40% in mediastinal lymph nodes with CT [1] as well the significant false-positive rate associated with PDG-PET depending on the clinical setting [2], it is essential to obtain tissue from appropriate lymph node stations when determining operability. Radiologic staging alone is not sufficient in most clinical settings.
Mediastinoscopy has been the gold standard [3] in sampling mediastinal lymph nodes, yet other methods also exist. The utility of bronchoscopic blind transbronchial needle aspiration (TBNA) [4–6] and endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) are well described [7–9] and may obviate the need for mediastinoscopy or thoracotomy [10–12].
Recently, endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) has been evaluated as a minimally invasive mediastinal staging modality. Similar to EUS-FNA, EBUS-TBNA consists of a flexible videobronchoscope equipped with a linear ultrasound device at the distal end that allows real-time needle aspiration of lymph nodes. Multiple studies have shown the utility of EUS-FNA [7–14], but the literature examining EBUS-TBNA is comparatively new [15–20]. The purpose of this study was to determine the sensitivity, specificity, and accuracy of EBUS-TBNA in undiagnosed intrathoracic lymphadenopathy compared with the gold standard of mediastinoscopy, when possible.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Bronchoscopy was performed in all cases with a flexible videobronchoscope equipped with a 6-MHz linear ultrasound probe (Olympus model XBF-UC160F-OL8; Olympus America, Inc, Center Valley, PA; Fig 1). All procedures were performed by three bronchoscopists experienced in EBUS-TBNA in a dedicated bronchoscopy laboratory with appropriate nursing and respiratory staff.
|
Biopsy Procedure
Biopsies were performed using a 22-guage needle specially designed for use with the ultrasound bronchoscope. Once cytologically positive results were obtained, no lower-stage nodal stations were sampled unless treatment was affected by specific nodal involvement (ie, need for external beam radiotherapy). In patients who were surgical candidates with a radiographically normal mediastinum, all lymph nodes visible by EBUS were sampled; however, not every nodal station examined had demonstrable lymph nodes by ultrasound examination. Adequacy of the lymph node biopsy was determined by the presence of lymphocytes on the cytology specimen.
Statistical Analysis
Statistical analysis was performed using SPSS 14.0 software (SPSS Inc, Chicago, IL) for Windows (Microsoft Inc, Redmond, WA). Sensitivity (TP/TP + FN), specificity (TN/TN + FP), positive predictive value (TP/TP + FP), negative predictive value (TN/TN + FN), and diagnostic accuracy (TP + TN/TP + TN + FP + FN) were calculated for the 113 cases of malignancy.
True-positives were defined as any lymph node biopsy specimen that yielded malignant cells by EBUS-TBNA, with or without surgical confirmation. True-negatives and false-negatives were defined as specimens determined to be adequate by the presence of lymphocytes without malignant cells confirmed by either (1) examination of surgical specimen after mediastinoscopy or lung resection or (2) radiologic evidence by a change in the CT or FDG-PET size, or metabolic activity at a minimum of 3 months showing either no evidence for (true-negative) or evidence for (false-negative), nodal metastatic disease. False-positives were defined as cytologic specimens with malignant cells at the time of EBUS-TBNA with negative results at time of mediastinoscopy or lung resection. There were 113 true-positives, 32 true-negatives, one false-negative, and no false-positives.
The negative predictive value of EBUS-TBNA was 97% (95% confidence interval [CI], 95% to 99%) and the positive predictive value was 100%. Sensitivity was 99.1% (95% CI, 97% to 100%), and overall diagnostic accuracy was 98.7%.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
|
|
EBUS-TBNA was used as a primary diagnostic and staging modality in patients with abnormal mediastinal lymph nodes by FDG-PET, CT, or both, that required tissue confirmation. As a primary diagnostic tool, EBUS-TBNA was sufficient to establish a diagnosis in 66 patients by lymph node biopsy at stations accessible by mediastinoscopy. These patients would have otherwise required mediastinoscopy for diagnosis. All 152 patients studied had undergone prior imaging with CT, 77 (50.6%) also had FDG-PET imaging, and 20 (13.2%) had undergone combined CT/PET imaging.
Compared with radiologic staging by CT, PET, and CT/PET, staging with EBUS-TBNA resulted in nodal status downstaging of 18 of 113 patients (16%) and upstaging in 11 patients (9.7%) with malignancy. Four of the 11 up-staged patients had NSCLC and went from radiologic local disease to locally advanced disease, as determined from the EBUS-TBNA results, and were excluded from surgical resection. In all 4 patients, a diagnosis and stage were simultaneously obtained from contralateral mediastinal lymph nodes, which were radiologically negative by CT, FDG-PET, or both. The remaining 7 up-staged patients had malignancy other than lung carcinoma, and EBUS-TBNA represented the discovery of a distant metastatic site. In only three of these cases was surgical resection of the primary lesion a consideration, and therefore, they were subsequently deemed inoperable. Two of these patients had esophageal cancer, and 1 had primary head and neck carcinoma.
Of the 18 down-staged patients, 7 had NSCLC, and 2 of these were able to undergo resection based on negative EBUS-TBNA results, all of which were confirmed by surgical pathology. No change in management resulted from the down staging in the remaining 5 lung cancer patients. In these 5 patients, the stage was IIIB by radiologic criteria and IIIA by EBUS-TBNA. These patients were treated with chemotherapy and external beam radiotherapy. Of the remaining 11 down-staged patients, 5 had metastatic disease to the mediastinum from a primary tumor other than lung excluded by a negative EBUS-TBNA and their primary tumor was resected as planned. The remaining 6 patients had mediastinal lymph node metastases excluded on the basis of a negative EBUS-TBNA but were found to have metastatic disease by another diagnostic modality, and therefore, treatment was not altered by EBUS-TBNA results.
Twenty patients with primary suspicion of lung cancer eventually underwent surgical resection. Two had mediastinoscopy, followed by thoracotomy, and the remainder had lymphadenectomy at the time of thoracotomy. Seven of the surgical patients (35%) had a normal mediastinum by FDG-PET, CT, or both. In all 7 patients, EBUS-TBNA confirmed the radiologic stage. The nodal status in 5 of the surgical patients (20%) was lowered as a result of EBUS-TBNA, with 2 (10%) being converted from nonsurgical to surgical candidates on the basis of the EBUS-TBNA findings. One patient had radiologic stage IIIB disease, but the N1 and N2 nodes sampled were reactive, giving a final pathologic stage of IIB. The other patient had hilar and subcarinal lymph nodes metabolically active by FDG-PET, but EBUS-TBNA sampling of the subcarinal node was negative. The hilar node was diagnostic for malignancy, and the patient underwent subsequent resection of IIB NSCLC. Only 1 of the 20 surgical patients was found to have microscopic N2 level nodal metastasis at a lymph node station that was negative by previous EBUS-TBNA, resulting in conversion from pathologic stage IIB to IIIA disease. No major complications were associated with this procedure.
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Noninvasive lung cancer staging includes radiologic studies to assess the size and metabolic activity of both the primary lesion and intrathoracic lymph nodes. CT is usually the initial step in radiologic staging of mediastinal lymph nodes but has unacceptably poor accuracy and is prone to interobserver variability [9, 13, 20–22]. Although FDG-PET has become increasingly used, it has not been shown to reduce the need for pathologic lymph node examination, including mediastinoscopy [15, 23–25], except in stage I disease, where mediastinoscopy was not cost-effective in patients with a negative FDG-PET scan result of the mediastinum [26]. Correlation with radiologic stage I disease, however, has been shown to be poor, with 38% of patients improperly clinically staged compared with pathologic examination [27]. The issue remains that proper staging requires tissue sampling of radiologically abnormal and, perhaps even radiologically normal, mediastinal lymph nodes [16].
Several methods are used for invasive staging of intrathoracic lymphadenopathy. Traditional or "blind" transbronchial needle aspiration (TBNA) of lymph nodes is usually accomplished with flexible videobronchoscopy using a 19-gauge to 22-gauge needle [4–6]. Multiple passes at a target site are made until diagnostic tissue is obtained. Yield may also be increased using a 19-gauge histology needle [5] as well as rapid on-site cytopathologic evaluation [10, 28]. The fact remains, however, that overall diagnostic yield is poor compared with mediastinoscopy, with a reported sensitivity of 76% (range, 14% to 100%) and a negative predictive value of 71% (range, 36% to 100%) [29, 30].
Gould and colleagues [31] conducted a meta-analysis of more than 500 prospective studies examining the utility of TBNA in mediastinal lymph node staging. Thirteen studies met their inclusion criteria, resulting in a pooled sensitivity of 39% and specificity of 98%. In this analysis, sensitivity was directly related to the prevalence of metastatic disease [31].
EUS-FNA is accomplished using a videoendoscope equipped with an ultrasound probe to examine mediastinal lymph nodes and abdominal structures, including adrenal glands. Lymph node stations accessible by EUS-FNA include levels 4L, 5, 7, 8, and 9 [7, 9]. EUS-FNA is more suited to nodes on the left side than on the right owing to the anatomic location of the esophagus [9, 13]. EUS-FNA has been shown to be a highly accurate and cost-effective technique in mediastinal lymph node staging, with a sensitivity of 88% (range, 87% to 100%) and negative predictive value of 77% (range, 68% to 100%) [29, 30]. EUS-FNA has also been shown to decrease the number of subsequent futile thoracotomies [11, 12] but may be underused [32].
Recently, EBUS-TBNA of lymph nodes has been introduced. EBUS-TBNA uses a videobronchoscope equipped with a linear ultrasound transducer probe that allows real-time sampling of lymph nodes from levels 2, 3, 4, 7, 10, and 11. Lymph node stations not accessible by EBUS-TBNA include levels 5, 6, 8, and 9. A 22-gauge cytology needle is fed through the working channel, and multiple passes are made until diagnostic tissue is obtained (see Fig 3). Rapid on site cytopathology examination is commonly used to increase diagnostic yield.
|
In one series of 502 patients, the range of lymph nodes accessed included levels 2R (n = 53), 2L (n = 40), 3 (n = 35), 4R (n = 86), 4L (n = 77), 7 (n = 127), 10R (n = 39), 10L (n = 43), 11R (n = 40), and 11L (n = 33). Lymphocytes were present (indicating adequacy of specimen) in 94.5% of cases, and the diagnosis was established from the lymph node biopsy specimen in 93.5% [35]. Some of these lymph node stations are frankly inaccessible by any other modality. EBUS-TBNA may also be combined with EUS-FNA in the same procedural setting with synergistic effects [13, 19].
One valid criticism of the minimally invasive techniques of blind TBNA, EUS-FNA, or EBUS-TBNA is that the patients who undergo these procedures are markedly different than those who undergo mediastinoscopy in that most patients undergoing the former have bulky disease, whereas the latter usually have single-station nodal disease or no discernible disease radiographically. Thus, comparison of test characteristics is unfair and biased towards minimally invasive techniques. Herth and colleagues [16], however, showed in 100 patients scheduled for surgery with a radiologically normal mediastinum on CT that EBUS-TBNA had a negative predictive value of 96.3% with surgical confirmation in 119 sampled lymph nodes [16]. They found cancer in the lymph nodes in 1 of every 6 patients. These findings compare favorably with mediastinoscopy.
The gold standard for mediastinal lymph node staging remains mediastinoscopy. The diagnostic accuracy is high and complications are rare [36, 37]. In fact, the most recent invasive staging guidelines recommend mediastinoscopy as the initial test of choice in mediastinal lymph node staging in all cases except those with bulky, radiologically unresectable disease and in patients with a radiologically negative mediastinum and a single lung mass [30]. However, current evidence indicates that mediastinoscopy may be underused in everyday surgical practice.
Smulder and colleagues [38] showed that only 40% of mediastinoscopies performed adhered to gold standard techniques. They also showed that mediastinal lymph node sampling was performed in only 50% of cases and estimated that 20% of thoracotomies could have been avoided by performing mediastinoscopy before thoracotomy or by adhering to gold standard surgical techniques. Their study found that of 39 cases with unexpected N2 disease at pathologic staging, 16 were accessible by mediastinoscopy.
Little and colleagues [39] showed in 11,668 patients that practice patterns vary considerably, with mediastinoscopy performed in only in approximately one-quarter (3162) of surgical candidates and diagnostic lymphoid tissue obtained in one-third (1054) [39]. These findings are disturbing, because appropriate staging—specifically, increased use of mediastinoscopy—should be encouraged.
EBUS-TBNA is an additional tool to stage the mediastinum, which could potentially increase the number of patients with lung cancer who are appropriately staged. EBUS-TBNA as well as EUS-FNA have been examined in conjunction with mediastinoscopy and have been shown to prevent futile thoracotomy in patients with NSCLC [10, 16, 40].
EBUS-TBNA does have several limitations. Additional time, experience, and resources are required to learn the technique. Practitioners familiar with traditional TBNA techniques should find that familiarization with the appearance of mediastinal structures including lymph nodes requires the most training. We found that surgeons intimately familiar with mediastinal anatomy and trained in bronchoscopic techniques quickly gain expertise in this aspect of the procedure. In our center, all EBUS-TBNA procedures are performed in the outpatient setting, under local anesthesia, with rapid on-site cytopathologic evaluation, which has been shown to improve the yield of samples obtained with blind TBNA [31], but this service may be unavailable in many centers.
Other limitations of EBUS-TBNA include cost and repair of equipment, limited field of view from the camera, and limited bronchoscope flexibility that make airway intubation and inspection slightly more difficult. EBUS-TBNA also requires additional training to become proficient.
EBUS-TBNA, however, has distinct advantages over mediastinoscopy. Although both mediastinoscopy and EBUS-TBNA are generally performed as an outpatient procedure and have very low associated morbidity and mortality, EBUS-TBNA avoids the risk of general anesthesia and surgical complications specific to mediastinoscopy, including nerve and vascular injury and infection [32]. Complications with TBNA are so rare that they are considered reportable events. In the eight studies of EBUS-TBNA reviewed here [15–19, 33–35], as well our series, no serious adverse events were associated with the procedure in a combined 629 patients. Mediastinoscopy has a reported morbidity of 2% to 3% and mortality of 0.0% to 0.8% [29, 30].
EBUS-TBNA can also access several lymph node stations either difficult or impossible to access by routine cervical mediastinoscopy, specifically the posterior subcarinal (level 7) and hilar stations (level 10 and above). When EBUS is combined with EUS-FNA, the range of nodal stations amenable to needle biopsy is extended to include all stations except the prevascular (level 6) lymph nodes and, in our experience, subcentimeter lymph nodes even as small as 4 mm could be reliably sampled with satisfactory results (lymphocytes or malignant cells, or both, present).
Real-time EBUS gives immediate visual feedback of proper needle placement that is lacking with conventional TBNA (see Fig 3). This attribute could theoretically reduce vascular complications associated with TBNA. In our experience, the tip of the biopsy needle is easily visualized and allows the bronchoscopist to stop the needle short of vascular structures.
Our study had several limitations. In only 20 cases were we able to evaluate surgically obtained tissue to confirm our bronchoscopic diagnoses. It would have been unethical to require surgical lymphadenectomy in all cases because many of our patients were deemed inoperable by radiologic criteria or by EBUS/TBNA.
Another limitation of EBUS-TBNA is the potential for false-positive results due to contamination from cytologically positive areas, including lymph nodes and the tracheobronchial tree. This was controlled for by sampling nodes beginning at the highest possible station in relation to the primary tumor (if present), and biopsies progressed in sequence from higher-stage to lower-stage nodes. In the event a false-positive result did occur, the tendency would be to up-stage to a lower nodal level.
The 22-gauge diameter of the biopsy needle is another potential limitation. In fact, the inner diameter of the needle is closer to that of a 21-gauge needle. We found core samples were not uncommon with this needle, especially in lymphadenopathy due to sarcoidosis.
Finally, in not every case where EBUS-TBNA results were either nondiagnostic or negative and surgical lymphadenectomy was not performed were we able to completely exclude the possibility of false-negative EBUS-TBNA results. We decided that a minimum of 3 months of clinical follow-up gave us enough time to realize potential malignant disease. Only eight such cases occurred in our series and in no case has there been a diagnosis of malignancy after a range of follow-up from 3 to 10 months.
In summary, EBUS-TBNA is a safe and effective means of sampling mediastinal nodes and accurately staging the mediastinum in patients with NSCLC. Further studies with direct comparison with mediastinoscopy should be performed to confirm the diagnostic accuracy of EBUS-TBNA. This tool can be a useful first test when a patient presents with undiagnosed mediastinal lymphadenopathy, with or without primary lung mass, because diagnosis and staging can both be accomplished simultaneously. In addition, for those patients with stage IIIA lung cancer, initial staging with EBUS-TBNA before neoadjuvant chemotherapy may avoid the problem of mediastinal scarring caused by an initial staging mediastinoscopy.
It is our opinion that EBUS will become more widely available to pulmonologists and thoracic surgeons. The high yield, the ability to perform the procedure under conscious sedation, the low complication rate, and perhaps more widespread acceptance of the technology make this an attractive procedure.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A.R.L. Medford, S. Agrawal, C.M. Free, and J.A. Bennett A performance and theoretical cost analysis of endobronchial ultrasound-guided transbronchial needle aspiration in a UK tertiary respiratory centre QJM, September 29, 2009; (2009) hcp136v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Gilbert, D. O. Wilson, N. A. Christie, A. Pennathur, J. D. Luketich, R. J. Landreneau, J. M. Close, and M. J. Schuchert Endobronchial ultrasound as a diagnostic tool in patients with mediastinal lymphadenopathy. Ann. Thorac. Surg., September 1, 2009; 88(3): 896 - 900. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. A. Silvestri The Mounting Evidence for Endobronchial Ultrasound Chest, August 1, 2009; 136(2): 327 - 328. [Full Text] [PDF]
|
|
|
|
 |
|
 L. Varela-Lema, A. Fernandez-Villar, and A. Ruano-Ravina Effectiveness and safety of endobronchial ultrasound-transbronchial needle aspiration: a systematic review Eur. Respir. J., May 1, 2009; 33(5): 1156 - 1164. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gomez and G. A. Silvestri Endobronchial Ultrasound for the Diagnosis and Staging of Lung Cancer Proceedings of the ATS, April 15, 2009; 6(2): 180 - 186. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Melloni, M. Casiraghi, A. Bandiera, P. Ciriaco, A. Carretta, L. Libretti, and P. Zannini Transbronchial needle aspiration in lung cancer patients suitable for operation with positive mediastinal positron emission tomography. Ann. Thorac. Surg., February 1, 2009; 87(2): 373 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Chen and Y.-m. Zhou Extended Mediastinoscopic Examination at the Right Hilum Ann. Thorac. Surg., November 1, 2008; 86(5): 1704 - 1706. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. S. Groth, B. A. Whitson, J. D'Cunha, M. A. Maddaus, M. Alsharif, and R. S. Andrade Endobronchial Ultrasound-Guided Fine-Needle Aspiration of Mediastinal Lymph Nodes: A Single Institution's Early Learning Curve Ann. Thorac. Surg., October 1, 2008; 86(4): 1104 - 1110. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. R. M. Abdel Rahman, R. M. Gaafar, H. A. Baki, H. M. El Hosieny, F. Aboulkasem, E. G. Farahat, A. M. Nouh, and K. A. Mansour Prevalence and Pattern of Lymph Node Metastasis in Malignant Pleural Mesothelioma Ann. Thorac. Surg., August 1, 2008; 86(2): 391 - 395. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
Diagnoses represent resected lung nodules, not lymph nodes.