© 2008 American Thoracic Society doi: 10.1164/rccm.200710-1605ED
A Good Case for a Declining Role for Mediastinoscopy Just Got Better
Beth Israel Deaconess Medical Center The discovery of metastatic N2 or N3 disease has significant implications for the prognosis and optimal care of patients with lung cancer. Positron emission tomography (PET) and PET–computed tomography (PET-CT) scanning have advanced the accuracy of clinical staging over CT scan alone (1). However, when evaluated prospectively, the combination of CT and PET-CT still remains relatively unsatisfactory, and unsuspected N2 disease may be proven pathologically in up to 9% of patients with clinical stage I and 26% of patients with clinical stage II disease (2). Mediastinoscopy has been the mainstay of prethoracotomy pathological staging for patients with lung cancer. However, it is not without its drawbacks, both in theory and in practice. Mediastinoscopy is a procedure that requires general anesthesia, and while its complication rate is around 1%, severe hemorrhage can occur in nearly 0.3% of cases, requiring emergent sternotomy or thoracotomy (3–5). Mediastinoscopy can sample only a fraction of the mediastinal lymph node stations. Standard cervical mediastinoscopy can access the paratracheal and subcarinal lymph node stations (2R, 2L, 4R, 4L, 7), but not the paraesophageal, inferior pulmonary ligament, and aortopulmonary window lymph node (stations 8, 9 5, 6). In addition, the lower aspect of the subcarinal station may be inaccessible via mediastinoscopy. Not surprisingly, the vast number of N2 nodes missed with mediastinoscopic staging tend to be found in those latter stations (3, 4). Similar to any operator-dependent procedure, there is variability in how effectively mediastinoscopy is actually performed. In a survey of practice patterns of 729 U.S. hospitals in which over 40,000 patients received surgical care for lung cancer in 2001, only 28% underwent mediastinoscopy prior to surgical resection (6). Keeping in mind that these data predate widespread usage of PET-CT scanning, the strikingly low proportion implies that a number of patients may have been understaged and thus did not receive optimal care. Even more concerning is the performance metric of nodal tissue obtained in the fraction of patients who underwent mediastinoscopy, as only 47% of these patients were found to have documentation of lymph node specimens submitted for pathological examination. Although the last two large (>2,000 patients each), single-institution series of mediastinoscopy report sensitivity rates around 85% (3, 4), the actual accuracy of mediastinoscopy as practiced nationally appears to be more suspect. It is therefore not surprising that minimally invasive technologies are garnering increasing interest in the staging of lung cancer. Endoscopic ultrasound (EUS)–guided aspiration through the esophageal route and endobronchial ultrasound (EBUS)–guided needle aspirations have received increased attention. In the article by Tournoy and colleagues (7), which appears in this issue of the Journal (pp. 531–535), a most important first step is taken: the "gold standard" of mediastinoscopy is compared in a randomized fashion with EUS-guided fine needle aspiration (FNA) as a first-step staging procedure. In a cohort of 40 patients with proven or suspected non–small cell lung cancer in need of staging, subjects were randomly assigned to undergo either primary mediastinoscopy or EUS-FNA. All negative EUS-FNA results were surgically controlled and follow-up data were available for all patients. Endpoints chosen were the rate of surgical interventions, as well as test performance and economic impact in terms of hospital stay for the procedure. In this methodologically well-designed trial, 32% of patients in the EUS-FNA group required further surgical staging and no patient had any significant complication. In the mediastinoscopy group, there was one severe complication (5%) of esophageal perforation, which has rarely been reported in the literature (8). The authors conclude that EUS-FNA should be considered to be the first step in patients in need of mediastinal staging with suspected mediastinal nodal disease as suggested by CT or PET-CT. The authors acknowledge several shortcomings: fewer stations are staged with EUS-FNA than with mediastinoscopy, because reaching the 4R nodes can be difficult. Also, surgical confirmation of N3 disease was not done, and given the median of a single nodal station sampled, it is possible that more advanced disease was missed in the EUS cohort. This may be the most significant issue related to all targeted endoscopic staging. Complete and reliable pathological staging is important not just for decision making but also for appropriate listing in studies and cancer registries. In addition, a more comprehensive interpretation of the results would be possible if the authors had provided additional detail on the target nodal size and location. This is significant because the study only included patients with suspicious mediastinal nodes, but a pathologically enlarged node is a different target for FNA than a solely fluoro-deoxyglucose (FDG)-avid lesion. In addition, only one patient underwent anterior mediastinotomy, and there is a suggestion that EUS may not be ideal for level 5 and 6 metastases (9). Last, mediastinoscopy was performed on an inpatient basis with a median stay of 2 days, skewing the results of the economic analysis, as the procedure can safely be performed on an outpatient basis. The advantage of any endoscopic staging procedure is that far more lymph node stations can be reached as compared with mediastinoscopy. EUS- and EBUS-guided needle aspirations may be complementary: EUS affords access to celiac lymph nodes and the left adrenal, EBUS can reach the hilar nodes and nodes anterior or to the right of the trachea—areas difficult to reach with the EUS instrument. In fact, there is evidence that combining EUS- and EBUS-guided staging procedures may actually provide the most comprehensive staging currently available (10). In addition, endoscopic staging can be repeated as needed. Lung cancer staging is in flux. Additional well-designed studies, such as the one performed by Tournoy and colleagues, are needed to establish the best staging strategies clinically and economically. These randomized and well-controlled comparisons should include EUS-FNA, EBUS-TBNA (transbronchial needle aspiration), and mediastinoscopy as sole interventions and in combination. The old adage that "tissue is the issue" is still holding true and mediastinoscopy will always be an important tool in staging the mediastinum, but the days of mediastinoscopy as the first-step procedure of choice in all patients to obtain it may well be numbered. FOOTNOTES Conflict of Interest Statement: A.E. has no direct conflict of interest to report; Olympus America, a manufacturer of EUS and EBUS equipment has supported CME courses at Harvard Medical School with unrestricted grants. S.P.G. has no financial relationship with a commercial entity that has an interest in the subject of this manuscript. REFERENCES
Related articles in AJRCCM:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
