© 2008 American Thoracic Society
A Guide to Staging in Lung Cancer, but Potentially Misleading without Attention to the DetailsTo the Editor:The study by Dr. Tournoy and colleagues (1) is quite unique as a randomized comparison of mediastinal staging tests in lung cancer, but details of the outcomes and the patients included are important to appropriately interpret the results. The design of the study is inherently biased in favor of endoscopic ultrasound (EUS), because the major endpoint (the need for surgical staging) can only be improved by EUS. While the goal of avoiding mediastinoscopy is certainly desirable, there is bias because the cost of a second procedure (i.e., negative EUS followed by mediastinoscopy) is not considered. The comparison of hospital stay is regionally dependent. A median hospital stay of 2 nights for mediastinoscopy is unthinkable in the United States, where this is routinely an outpatient procedure. Furthermore, esophageal perforation during mediastinoscopy is extremely rare, and must be viewed as an outlier. The characteristics of included patients are not well defined, so readers may apply these data to inappropriate patient cohorts. The key issues are the location and size of suspicious nodes because these affect the performance characteristics of staging tests (2). The study appears to have involved primarily patients with bulky mediastinal node enlargement. First, all patients had suspected N2,3 node involvement, although this could be due to positron emission tomography (PET) uptake in a normal-sized node. Second, 90% of the patients in the EUS first arm had at least one enlarged node. Finally, the study would be impractical unless there was diffuse mediastinal node enlargement. It would be awkward to randomize patients with only one suspicious node between staging tests that may not allow access to the node in question. In many patients with diffuse node enlargement, the issue is not to confirm the mediastinal stage but to define the type of lung cancer (2). It would have been a cleaner and better defined study if patients undergoing biopsy only to confirm the diagnosis had been excluded. The performance characteristics reported in Table 5 are potentially misleading because most readers will assume the results represent EUS alone. It is unclear from both the table and the text whether the EUS results pertain to EUS alone, or to the approach of EUS first, followed by mediastinoscopy if negative (both of which are of interest). Despite these ambiguities and minor flaws, this is an important study and a welcome addition to our knowledge base. The authors should be commended for conducting this research.
Yale University FOOTNOTES Conflict of Interest Statement: F.D. has no financial relationship with a commercial entity that has an interest in the subject of this manuscript. REFERENCES
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
