In 1982 Pearson and colleagues [1] from Toronto published

Transcription

1 Transition From Mediastinoscopy to Endoscopic Ultrasound for Lung Cancer Staging Mark I. Block, MD Division of Thoracic Surgery, Memorial Healthcare System, Hollywood, Florida Background. Esophageal endoscopic ultrasound (EUS) and endobronchial ultrasound (EBUS) are gaining popularity for mediastinal staging of patients with lung cancer. Endoscopic ultrasound and then EBUS were introduced into a single-surgeon thoracic surgical practice. Records were reviewed to determine what effect this had on performance of mediastinoscopy for lung cancer staging, and on discovery of unsuspected N2 disease at the time of resection. Methods. Endoscopic ultrasound and EBUS were introduced 10 months apart. Records were reviewed for the 10 months before EUS (phase 1), the 10 months between EUS and EBUS (phase 2), 8 months after the introduction of EBUS (phase 3), and 11 months after that (phase 4). The number of staging procedures, patients undergoing resection after negative staging, and patients with N2 disease discovered at resection were determined. Results. Two hundred fifty-three patients met inclusion criteria. Eighty-two had resection without staging; staging was positive in 62, negative in 90 who went on to resection, and negative in 19 who had no further evaluation. There was a strong trend toward preferential use of EUS in phase 2 and EBUS in phases 3 and 4. Nine patients (10%) had N2 disease found at surgery: 0 of 16 in phase 1, 4 of 24 in phase 2, 3 of 24 in phase 3, and 2 of 25 in phase 4. Overall sensitivity and negative predictive value of EUS and EBUS were 84% and 87%, respectively. Conclusions. Introduction of EUS and EBUS reduced use of mediastinoscopy. Discovery of N2 disease at surgery was higher than expected initially, but with experience results proved comparable to those of mediastinoscopy. (Ann Thorac Surg 2010;89:885 90) 2010 by The Society of Thoracic Surgeons In 1982 Pearson and colleagues [1] from Toronto published their results with the use of mediastinoscopy for preoperative evaluation of patients with lung cancer. They found that actuarial survival after lung cancer resection was worse for patients with N2 disease that was detected at mediastinoscopy than for patients who had a negative mediastinoscopy but N2 disease discovered at resection. Since then mediastinoscopy has been the accepted standard for mediastinal staging. Recent technologic advances have led to the development of endoscopic procedures for real-time ultrasound-guided fine-needle aspiration of mediastinal lymph nodes. Transesophageal endoscopic ultrasound-guided (EUS) fine-needle aspiration was introduced more than 10 years ago [2]. Endobronchial ultrasound (EBUS) -guided transbronchial biopsy was introduced in 2004 [3]. Endoscopic ultrasound and EBUS afford access to multiple mediastinal lymph node stations. Endoscopic ultrasound enables biopsy of the left paratracheal, subcarinal, and paraesophageal lymph nodes as well as both adrenal glands. Endobronchial ultrasound provides access to bilateral paratracheal, subcarinal, and bilateral hilar lymph nodes. Supported by advancing expertise in cytologic analysis of needle aspiration specimens, EUS and Accepted for publication Nov 12, Presented at the Fifty-fifth Annual Meeting of the Southern Thoracic Surgical Association, Austin, TX, Nov 5 8, Address correspondence to Dr Block, Division of Thoracic Surgery, Memorial Healthcare System, 1150 N 35th Ave, Ste 660, Hollywood, FL 33021; mblock@mhs.net. EBUS have attracted attention as minimally invasive but potentially reliable techniques for complete assessment of the mediastinum. Published data indicate that the sensitivity of EUS and EBUS for detection of mediastinal metastases is approximately 90% [4 8]. For the most part these reports have come from a small number of institutions and investigators who have been at the forefront of development of these technologies. There is little to no data on results from the use of EUS and EBUS by individuals and at institutions that are not early adopters. This study was undertaken to examine the experience of a community thoracic surgical practice with transition from mediastinoscopy to EUS for mediastinal staging. Endoscopic ultrasound was introduced into the practice in June 2006, and EBUS was introduced 10 months later in April This produced a phased transition from exclusive use of mediastinoscopy to dominant use of EBUS. The focus of this review was to determine whether this migration led to a change in the quality of care provided as measured by a change in the frequency with which patients who had had negative invasive mediastinal staging were found to have mediastinal disease at operation the false-negative rate. Material and Methods Institutional Review Board The Institutional Review Board of the Memorial Healthcare System, Hollywood, FL, approved this retrospective 2010 by The Society of Thoracic Surgeons /10/$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 886 BLOCK Ann Thorac Surg ENDOSCOPIC ULTRASOUND FOR LUNG CANCER STAGING 2010;89: review. The requirement for informed consent was waived. Invasive Mediastinal Staging All invasive mediastinal staging was performed as a part of routine clinical care for patients with known or suspected lung cancer. All patients had chest computed tomography (CT) before the procedure, and most had dedicated positron emission tomography (PET) or PET/CT studies available. In general American College of Chest Physicians guidelines for invasive mediastinal staging were followed [9]. Thus, sampling was performed routinely in patients with evidence of mediastinal disease by imaging (class B) or with central tumors or evidence of hilar disease (class C). Sampling was not performed routinely in patients with small, peripheral tumors and no evidence of mediastinal disease by imaging (class D). The decision regarding which procedure(s) to recommend was based on procedure availability and on clinical features such as location of the primary lung tumor, location of abnormal mediastinal lymph nodes on CT and PET, and patient characteristics such as prior mediastinoscopy and limited cervical extension. All patients were evaluated by a single surgeon, and all procedures were performed by the same surgeon. Mediastinoscopy Standard cervical mediastinoscopy was performed in all cases in which mediastinoscopy was done. The procedure was done in the operating room under general anesthesia either as a separate procedure (outpatient) or immediately before a planned resection. Endoscopic Ultrasound All endoscopic ultrasound procedures (EUS and EBUS) were done in the endoscopy suite on spontaneously breathing patients under monitored anesthesia care (intravenous sedation, typically with propofol, fentanyl, or midazolam). Patients undergoing EBUS also received nebulized lidocaine immediately before the procedure and topical lidocaine during the procedure. All elective procedures were done on an outpatient basis. Esophagogastroduodenoscopy was always performed immediately before EUS, and flexible bronchoscopy was always performed immediately before EBUS. If a combined EUS and EBUS were planned for the same patient, EBUS was usually performed first to take advantage of the nebulized lidocaine. Endoscopic ultrasound was performed first only if the patient had suspicious findings in the left adrenal gland on imaging studies. Endoscopic ultrasound fine-needle aspiration was performed using a flexible esophagoscope equipped with a linear ultrasound array (model GF-UCT140; Olympus Ltd, Tokyo, Japan). Endobronchial ultrasound-guided transbronchial biopsy was performed using a video bronchoscope equipped with a linear ultrasound array (model XBF-UC160F-OL8; Olympus Ltd, Tokyo, Japan). An attempt was made to biopsy all visualized lymph nodes. Biopsy specimens were taken first from contralateral (N3) mediastinal lymph nodes, followed by ipsilateral (N2) and then hilar (N1) lymph nodes. Multiple passes were made at each lymph node, and the needle was flushed with normal saline solution between subsequent stations. Specimens were spread onto glass slides and immediately fixed in alcohol for cytologic examination. Occasionally, material from multiple aspirations of the same station was collected in RPMI media for processing into a cell block to facilitate performance of special stains. Rapid on-site evaluation (ROSE) was not available; however, occasional samples were sent immediately to pathology for a STAT reading. Most samples were processed routinely with pathology results available in 1 to 3 days. Ultrasound-guided biopsy results were categorized as positive if malignant cells were identified in the specimen. Samples either negative (abundance of normal lymphocytes without diagnostic malignancy, consistent with lymph node assessment), nondiagnostic (insufficient lymphocytes to indicate reliable lymph node assessment), or atypical but not clearly malignant were categorized as negative. Resection All resections were done through either a posterolateral thoracotomy or video-assisted thoracoscopic surgery approach. Mediastinal lymph node sampling or dissection was performed routinely. For patients evaluated with endoscopic ultrasound, resection was at least 3 days later. Study Period Division Into Four s Endoscopic ultrasound and EBUS were introduced in June 2006 and April 2007, respectively. Therefore the study period was divided into phases that represented roughly equal periods of time before and after introduction of each new modality. 1, the 10 months before introduction of EUS, extended from August 1, 2005, to June 1, , the 10 months during which EUS was available but before introduction of EBUS, extended from June 1, 2006, to April 1, extended from April 1, 2007, to December 1, 2007, and was the 8 months after introduction of EBUS when all three invasive stag- Table 1. Invasive Mediastinal Staging Procedures Performed a Procedure Total Mediastinoscopy EUS EBUS EUS EBUS US Mediastinoscopy All a The number of patients undergoing each type of invasive staging procedure during each phase of the study is shown. EUS EBUS indicates both procedures were performed at the same time on the same patient. US Mediastinoscopy indicates that a mediastinoscopy was done after a negative EUS or EBUS. EBUS endobronchial ultrasound; EUS esophageal endoscopic ultrasound; US ultrasound.

3 Ann Thorac Surg BLOCK 2010;89: ENDOSCOPIC ULTRASOUND FOR LUNG CANCER STAGING 887 Table 2. Incidence of Positive Invasive Mediastinal Staging a Procedure Cumulative Mediastinoscopy 9/26 (35%) 4/15 (27%) 2/5 (40%) 6/9 (67%) 21/55 (38%) EUS... 9/29 (31%) 6/14 (43%) 7/18 (39%) 22/61 (36%) EBUS /35 (31%) 13/44 (30%) 24/79 (30%) All 9/26 (35%) 13/44 (30%) 19/54 (35%) 26/71 (37%) 67/195 (34%) a The denominator indicates the total number of staging procedures performed. The numerator indicates the number of those procedures that were positive for mediastinal (N2 or N3) disease. EBUS endobronchial ultrasound; EUS esophageal endoscopic ultrasound. ing modalities were available. Preliminary results of this study were submitted in abstract form based on these three phases. The additional time from abstract submission to presentation of the data permitted addition of a fourth phase, extending from December 1, 2007, to October 1, thus was an additional 11-month period when all three modalities were available and when results could reasonably be expected to represent their mature application with minimal impact from a learning curve. Patients were assigned to one of the four phases depending on the date invasive mediastinal staging was performed. Results Study Population Two hundred fifty-three patients with lung cancer were evaluated (236 non small cell lung cancer, 4 atypical carcinoid, 4 bronchoalveolar cell carcinoma, 4 large cell carcinoma, 4 small cell carcinoma, and 1 carcinosarcoma). Of these, 82 patients underwent resection without preoperative invasive mediastinal staging, whereas 171 patients had one or more invasive mediastinal staging procedures performed (Table 1). Mediastinal lymph nodes were abnormal by imaging in 75 (44%) of these 171 patients (46%, 37%, 51%, and 42%, respectively for phases 1 through 4). Forty-one patients (24%) had American College of Chest Physicians class D disease (35%, 24%, 23%, and 19%, respectively for phases 1 through 4). Only one invasive modality was used for most patients, but 18 patients had a combined EUS and EBUS approach, 3 patients had a negative EUS followed by mediastinoscopy, and 3 patients had a negative EBUS followed by mediastinoscopy. Lymph Node Sampling The number of mediastinal lymph node stations sampled is a surrogate marker for the completeness of the staging procedure, and could correlate with the sensitivity of the procedure for detection of metastatic disease. The average number of stations sampled at mediastinoscopy was relatively stable at approximately 2.5 through the course of the entire study period (2.8, 2.2, 2.6, and 2.3 for phases 1 through 4, respectively). In contrast, the average number of stations sampled at EUS rose from only 1.8 in phase 2, to 2.4 in phase 3, and finally to 3.1 in phase 4. Similarly, the average number of stations sampled at EBUS increased from 2.5 in phase 3 to 3.2 in phase 4. Results of Invasive Staging Procedures Invasive mediastinal staging was positive in 62 of the 171 patients (36%). None of these 62 patients went on to immediate resection. Because several patients had more than one procedure performed, the total number of positive procedures was 67 of 195 performed (34%). Detailed results are shown in Table 2, and demonstrate a higher yield with mediastinoscopy compared with EUS and EBUS, particularly in phase 4. This probably reflects a selection bias. Throughout the study period EUS and EBUS equipment was available at only one of the two major hospitals in the health-care system. So although EUS and EBUS became the preferred technique for elective evaluation of outpatients, patients admitted to the hospital that did not have EUS and EBUS were more likely to undergo mediastinoscopy. Because inpatients are more likely to have advanced disease, this created a bias toward performing mediastinoscopy on patients with advanced disease such that in phase 4, 5 of the 6 patients undergoing mediastinoscopy had mediastinal adenopathy by CT. Findings at Resection Of the 109 patients with negative invasive staging, 90 went on to resection. Eighty-one patients (90%) had an Table 3. Incidence of N2 Disease Discovered at Surgery Procedure Cumulative Mediastinoscopy 0/17 1/8 0/3 0/1 1/29 EUS... 2/14 0/2 0/1 2/17 EBUS /13 0/14 1/27 EUS EBUS /6 1/7 3/13 EUS Med... 1/ /2 EBUS Med /2 1/2 All 0/17 4/24 3/24 2/25 9/90 a The denominator is the total number of patients operated on for whom the indicated preoperative staging procedure was negative. The numerator is the number of those patients found at surgery to have N2 disease. EBUS endobronchial ultrasound; EUS esophageal endoscopic ultrasound; Med mediastinoscopy.

4 888 BLOCK Ann Thorac Surg ENDOSCOPIC ULTRASOUND FOR LUNG CANCER STAGING 2010;89: Table 4. Sensitivity and Negative Predictive Value of Mediastinal Staging a Mediastinoscopy, EUS, and EBUS EUS and EBUS only TP TN FN Sensitivity NPV TP TN FN Sensitivity NPV % 100% % 83% % 81% % 88% % 86% % 92% % 92% All % 90% % 87% a The number of patients with positive staging procedures (TP), negative staging procedures confirmed at surgery (TN), and negative staging procedures but with N2 disease discovered at surgery (FN) are shown for all staging procedures (mediastinoscopy, EUS, and EBUS), and for only ultrasound-guided procedures (EUS and EBUS). Sensitivity and negative predictive value for detection of mediastinal disease were calculated as TP/(TP FN) and TN/(TN FN), respectively. EBUS endobronchial ultrasound; EUS esophageal endoscopic ultrasound; FN false negative; NPV negative predictive value; TN true negative; TP true positive. anatomic resection (6 pneumonectomy, 72 lobectomy, and 3 segmentectomy). The average number of lymph node stations sampled at the time of resection was 2.6, 1.7, 2.3, and 2.8 in phases 1 through 4, respectively. Nine patients (10%) had N2 disease discovered at surgery (Table 3). In phase 2, the false-negative rate was relatively high at 17% (4 of 24 patients). However, in phases 3 and 4 the rate declined to 12.5% (3 of 24 patients) and then 8% (2 of 25 patients). Of the 4 patients in phase 2 with missed N2 disease, 1 had a mediastinoscopy, 1 had a negative EUS followed by a negative mediastinoscopy, and the remaining 2 had negative EUS. In phase 3, 2 of the 3 patients with missed N2 disease had a combined EUS and EBUS staging procedure, whereas in phase 4 1 of the 2 patients with missed N2 disease also had a combined EUS and EBUS and the other had a negative EBUS followed by a negative mediastinoscopy. Thus, for 5 of the 9 patients with missed N2 disease, two modalities were used for invasive mediastinal staging. This suggests a high clinical suspicion in these patients. Performance of Invasive Mediastinal Staging Table 4 shows the calculated sensitivities and negative predictive values (NPV) of invasive staging during the four phases of this study. Overall, the sensitivity for detection of mediastinal disease was 87%, whereas for EUS and EBUS excluding mediastinoscopy, it was 84%. Similarly the NPV overall was 90%, whereas for EUS and EBUS excluding mediastinoscopy it was 87%. Both the sensitivity and NPV of EUS and EBUS increased steadily with time, suggesting not surprisingly that increasing experience with the techniques and their clinical application led to better results. Calculations of specificity and positive predictive value were not performed because the specificity of detecting malignant disease in this setting is generally considered to be 100%, and because no patients with a positive staging procedure underwent confirmatory surgical biopsy. Clinical Course of Patients With False-Negative Staging Of the 9 patients with false-negative invasive mediastinal staging, 2 died of their disease shortly after surgery, and 3 have progressive disease. The remaining 4 patients are disease-free with short follow-up. Three of the 4 patients in phase 2 and 1 of the 3 patients in phase 3 had CT or PET findings suspicious for aortopulmonary (AP) window disease. All 4 had disease confirmed at this site at surgery. This suggests that EUS and EBUS were unable to accurately assess AP window nodes. Comment Because the purpose of invasive mediastinal staging is to identify stage III disease before resection, the goal of this study was to determine whether transition from mediastinoscopy to EUS and EBUS led to a higher-thanexpected rate of N2 disease discovered at resection. Although the false-negative rate may be the most conspicuous failure of the staging procedure(s), it is highly susceptible to changes in the pretest probability of disease. Assuming the performance characteristics of a test remain constant, an increase in pretest probability produces a higher false-negative rate and decreases the NPV. In contrast, sensitivity is less directly dependent on pretest probability. Therefore, for comparison of findings from one study of mediastinal staging to another, sensitivity may be the best yardstick. In this study, the early experience with introduction of EUS produced a sensitivity for detection of mediastinal disease of only 73%. However, with continuing experience and addition of EBUS, a steady improvement was seen. Sensitivity of EUS and EBUS rose to 90% in phase 4 and was 84% overall. This compares favorably with other studies of EUS and EBUS, which report sensitivities (on a per-patient not per-node basis) of 73% to 93% [4 8], and to the sensitivity of mediastinoscopy reported in the literature (85% to 85%) [7, 10 12]. The NPVs in these large series of mediastinoscopy were higher than in the current study (91% to 95% versus 87%), but the prevalences of disease were lower (29% to 39% versus 47%). This suggests that the lower NPV seen in this study may reflect a difference in the study populations and not in the reliability of EUS and EBUS compared with mediastinoscopy. The findings of this study suggest that introduction of EUS and EBUS, and transition from medias-

5 Ann Thorac Surg BLOCK 2010;89: ENDOSCOPIC ULTRASOUND FOR LUNG CANCER STAGING 889 tinoscopy to EBUS as the primary mode of invasive mediastinal staging, did not appreciably alter the ability to detect mediastinal disease preoperatively. Review of the nine false-negative cases suggests that surgical biopsy should be considered when imaging is positive but the EUS or EBUS is negative, particularly early in the experience. Seven of the 9 patients had abnormal preoperative imaging of the mediastinum, and 4 had metastatic disease to AP window lymph nodes. Endoscopic ultrasound especially has been promoted as capable of reaching the AP window. The experience in this study suggests otherwise. Typically, the AP window nodes are seen on the opposite side of the aorta from the transducer. The angle at which the needle is advanced from the endoscope, even with the elevator fully deflected, is usually not sufficiently acute to reach laterally through the gap between the aorta and pulmonary artery to nodes adjacent to the aorta (level 5). It is possible to reach nodes adjacent to the pulmonary artery, but for most AP window nodes a transaortic puncture is probably necessary. Although transaortic puncture may be safe from the standpoint of vascular injury, this maneuver necessitates traversing the aorta with a needle potentially coated with tumor cells. There is no evidence that this can increase the risk of hematogenous metastases, but it seems to tempt fate unnecessarily. The experience of the current study supports the conclusions of Cerfolio and colleagues [13] that video-assisted thoracoscopic surgery may be the best approach to access the AP window nodes. A strength of this study is that it is a relatively large series from a single surgeon at a community hospital. This creates consistency not only in the technical aspects of all three staging procedures, but also in the clinical decision making regarding which procedure to perform. It also addresses the concern expressed by Wigle [14] that [t]he data we have seen so far...[are] really by the innovators in endoscopic staging, and may not be representative of real world results. But simultaneously it may also hamper the generalizability of the findings because few centers will have just one physician making the clinical decisions and performing all the staging procedures and resections. As demonstrated in this report EUS and EBUS can be rapidly integrated into a thoracic surgical practice. Because a surgeon is experienced with endoscopy, invasive procedures, and mediastinal anatomy, formal training and proctoring although helpful may not be necessary. In this experience EUS was adopted after orienting to the equipment and observing experts. Once experience was gained with EUS, EBUS could then be adopted after a similar brief period of orienting to the equipment and training with a simulator. Becoming familiar with the ultrasound images and correlating them with mediastinal anatomy is the key to successful application of EUS and EBUS. For this reason EUS proved to be an excellent transitional technique. Learning the anatomy with EUS is easier because the transducer is larger and because air less often interferes with imaging. In addition there is much less resistance to needle puncture through the esophagus than through the airway. With EBUS, advancing the needle tends to push the airway away from the transducer and cause image loss. To overcome this resistance a sharp stabbing motion may be required. With EUS this is necessary much less often, making it easier to become rapidly comfortable with needle aspiration of specific targets. If undertaken in a deliberate fashion, the learning curve with adoption of EUS and EBUS should not adversely affect patients. Procedure-related morbidity should be very low, even in inexperienced hands, and mediastinoscopy should be considered if results of EUS or EBUS are negative in patients with suspicious imaging findings. Mounting evidence, including this report, suggest that EUS and EBUS provide equivalent if not improved mediastinal staging for patients with lung cancer. These results have been sufficient to foster introduction of these techniques into clinical practice at many academic and community centers. It is therefore likely that EUS and EBUS will supplant mediastinoscopy in most centers as the preferred method of mediastinal sampling. There are several reasons to expect this. First, although mediastinoscopy in experienced hands is safe and is associated with minimal morbidity, EUS and EBUS are endoscopic procedures with even less risk. The appeal to patients and referring physicians of an endoscopic approach over a surgical procedure that requires general anesthesia is undeniable. Second, there is an economic advantage to EUS and EBUS because it can be done in the endoscopy suite under sedation rather than in the operating room under general anesthesia. Third, necessary expertise with mediastinoscopy is not widely available. This reality was made startlingly clear by the findings of Little and colleagues [15]. Based on a survey of data on almost 41,000 patients nationwide, they reported that mediastinoscopy was performed in fewer than 30% of patients undergoing resection, and that, when performed, no lymph nodes were sampled in more than half the cases. Fourth, EUS and EBUS provide access to nodes that occasionally are unreachable by mediastinoscopy. This includes not only hilar and inferior mediastinal nodes, but also worrisome nodes in patients who have already had mediastinoscopy, resections with node dissections, and radiation therapy. And fifth, and perhaps most important, EUS and EBUS are accessible to nonsurgeons. With increasing interest in EUS and EBUS as alternatives to mediastinoscopy, it is critical to be mindful of the advantages of mediastinoscopy for lung cancer staging. Just as with limited availability of quality mediastinoscopy, there is likely to be limited availability of quality EUS and EBUS. Mediastinoscopy may also be preferred for those patients at low risk of mediastinal disease for whom resection is planned. In this setting mediastinoscopy done immediately before resection, under the same anesthetic, avoids a separate trip to the endoscopy suite, expedites care, and is probably less costly. The potential downside of transitioning from mediastinoscopy to EUS or EBUS is that mediastinoscopy may be eliminated altogether as a tool in the management of patients with lung cancer. As nonsurgeons become in-

6 890 BLOCK Ann Thorac Surg ENDOSCOPIC ULTRASOUND FOR LUNG CANCER STAGING 2010;89: creasingly capable of invasive mediastinal staging with EUS and EBUS, there may be a tendency to favor these techniques. Similarly, surgeons who have not incorporated EUS and EBUS into their practice may favor mediastinoscopy. The rapid adoption of EUS and EBUS for lung cancer staging requires acknowledging these inherent biases and guarding against their unduly affecting clinical recommendations. Maintaining a holistic approach to mediastinal staging is necessary to preserve the appropriate role of mediastinoscopy. One way to insure that this happens is for thoracic surgeons to integrate EUS and EBUS into their practice. References 1. Pearson FG, DeLarue NC, Ilves R, Todd TR, Cooper JD. Significance of positive superior mediastinal nodes identified at mediastinoscopy in patients with resectable cancer of the lung. J Thorac Cardiovasc Surg 1982;83: Silvestri GA, Hoffman BJ, Bhutani MS, et al. Endoscopic ultrasound with fine-needle aspiration in the diagnosis and staging of lung cancer. Ann Thorac Surg 1996;61: Yasufuku K, Chiyo M, Sekine Y, et al. Real-time endobronchial ultrasound-guided transbronchial needle aspiration of mediastinal and hilar lymph nodes. Chest 2004;126: Wallace MB, Silvestri GA, Sahai AV, et al. Endoscopic ultrasound-guided fine needle aspiration for staging patients with carcinoma of the lung. Ann Thorac Surg 2001;72: Wallace MB, Pascual JM, Raimondo M, et al. Minimally invasive endoscopic staging of suspected lung cancer. JAMA 2008;299: Eloubeidi MA, Cerfolio RJ, Chen VK, Desmond R, Syed S, Ojha B. Endoscopic ultrasound-guided fine needle aspiration of mediastinal lymph node in patients with suspected lung cancer after positron emission tomography and computed tomography scans. Ann Thorac Surg 2005;79: Annema JT, Versteegh MI, Veselic M, Voigt P, Rabe KF. Endoscopic ultrasound-guided fine-needle aspiration in the diagnosis and staging of lung cancer and its impact on surgical staging. J Clin Oncol 2005;23: Witte B, Neumeister W, Huertgen M. Does endoesophageal ultrasound-guided fine-needle aspiration replace mediastinoscopy in mediastinal staging of thoracic malignancies? Eur J Cardiothorac Surg 2008;33: Detterbeck FC, Jantz MA, Wallace M, Vansteenkiste J, Silvestri GA. Invasive mediastinal staging of lung cancer: ACCP evidence-based clinical practice guidelines (2nd edition). Chest 2007;132(Suppl):202S 20S. 10. Hammoud ZT, Anderson RC, Meyers BF, et al. The current role of mediastinoscopy in the evaluation of thoracic disease. J Thorac Cardiovasc Surg 1999;118: Lemaire A, Nikolic I, Petersen T, et al. Nine-year single center experience with cervical mediastinoscopy: complications and false negative rate. Ann Thorac Surg 2006;82: Luke WP, Pearson FG, Todd TR, Patterson GA, Cooper JD. Prospective evaluation of mediastinoscopy for assessment of carcinoma of the lung. J Thorac Cardiovasc Surg 1986;91: Cerfolio RJ, Bryant AS, Eloubeidi MA. Accessing the aortopulmonary window (#5) and the paraaortic (#6) lymph nodes in patients with non-small cell lung cancer. Ann Thorac Surg 2007;84: Wigle DA. The beginning of the end of mediastinoscopy? J Thorac Oncol 2008;3: Little AG, Rusch VW, Bonner JA, et al. Patterns of surgical care of lung cancer patients. Ann Thorac Surg 2005;80: