EUS staging of primary lung carcinoma: are we ready for it?

Abstract

The use of EUS in the staging and diagnosis of lung carcinoma has evolved over the past decade but not as rapidly nor as widely, compared with its GI applications. Since the first descriptions of EUS-guided FNA (EUS-FNA) in the setting of lung carcinoma,1.Wiersema M.J. Kochman M.L. Chak A. Cramer H.M. Kessler K. Real-time endoscopic ultrasound guided fine needle aspiration of a mediastinal lymph node.Gastrointest Endosc. 1993; 39: 429-431Abstract Full Text PDF PubMed Scopus (75) Google Scholar, 2.Wiersema M.J. Kochman M.L. Cramer H.M. Wiersema L.M. Preoperative staging of non-small cell lung cancer: transesophageal US-guided fine-needle aspiration biopsy of mediastinal lymph nodes.Radiology. 1994; 190: 239-242Crossref PubMed Scopus (52) Google Scholar a number of studies have been published describing the utility of EUS-FNA in the lymph nodal staging of non-small-cell lung carcinoma (NSCLC) and have found EUS to be cost-effective in the pre-resection staging evaluation for NSCLC. A few articles have noted primary tumor staging results, but none of these described a systematic approach with verification of the accuracy of the staging. The publication in this issue of Gastrointestinal Endoscopy by Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar is the first attempt to verify the accuracy of primary tumor staging of NSCLC when performed with EUS.In this retrospective study, performed by querying a database of 308 patients that contained pertinent information from the staging procedures that were performed at the Medical University of South Carolina for lymph nodal staging and tissue diagnosis, these investigators demonstrated that the accuracy of T4 staging was below that which traditionally would be expected for a staging modality. Because surgical resection remains the cornerstone of curative therapy for NSCLC, even with locally advanced disease, it is bothersome that inaccurate T staging by EUS may prohibit patients from undergoing a potentially curative surgical procedure. Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar provide data from a subset of their registry, with data gathered from 175 patients with surgical staging, 10 of whom were found to have T4 disease by EUS staging, and one in that cohort who was not staged as T4 but subsequently was demonstrated to have T4 disease at the time of surgery. For T4 lesions, these data yield a positive predictive value of 70%, a sensitivity of 87.5%, and a specificity of 98%. The specific patient population studied and the selection process used to refer the patients for the EUS staging procedure may significantly affect these numbers. Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar do not have surgical confirmation for the remaining 133 patients, of whom 5 were staged as having T4 disease by EUS; of these, 3 were staged by CT as having T3 cancer, and two were staged by CT as having T4 cancer.Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar referenced the old TNM staging and did not use the current staging schema,4.Greene F.L. Page D.L. Fleming I.D. Fritz A. Balch C.M. Haller D.G. et al.AJCC cancer staging manual.6th ed. Springer, New York2002Crossref Google Scholar but there is no difference in the T and N staging or the stage grouping of these two schemas. It is important to recognize and understand that bulky N2 or N3 lymph nodal disease, T4 primary tumors (other than intralobar satellite lesions), and distant metastatic disease generally preclude surgical cure, even in the setting of multimodality therapy.In addition, it is necessary to examine the clinical context in which the staging of lung carcinoma is occurring. At our institution, the majority of patients with suspected or proven NSCLC referred for mediastinal evaluation with EUS were felt to be poor surgical candidates or were highly suspected to harbor a postoperative recurrence. Why this institutional selection bias occurs is an important issue and pertinent to the options for the cost-effective evaluation and proper staging of patients with NSCLC. We work in an environment that has an active therapeutic endoscopy program, a busy interventional bronchoscopy service, and thoracic surgeons who are well versed and skilled at minimally invasive diagnostic and therapeutic procedures, which allows for a multidisciplinary approach. The staging of lung carcinoma in the pre-operative setting may be accomplished in any number of ways, some complementary and some competitive, but all share the ultimate goal of selecting those patients who will either clearly benefit from a surgical intervention and those who will clearly not benefit from intervention.A brief review of the available staging modalities is helpful in placing the study by Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar in the correct clinical context. The most widely disseminated modality to detect and stage a patient with lung cancer is helical CT. CT is invaluable for detecting distant metastasis, but the evaluation of regional lymph node metastasis is limited. A CT criterion for metastatic disease includes lymph node size of 1 cm or greater, which clearly limits its sensitivity. The range of sensitivity of CT in detecting mediastinal lymph nodes ranges from 43% to 89%, with a specificity of 32% to 94%.5.Silvestri G.A. Hoffman B.J. Bhutani M.S. Hawes R.H. Coppage L. Sanders-Cliette A. et al.Endoscopic ultrasound with fine needle aspiration in the diagnosis and staging of lung cancer.Ann Thorac Surg. 1996; 61: 1441-1446Abstract Full Text PDF PubMed Scopus (258) Google Scholar, 6.Lloyd C. Silvestri G.A. Mediastinal staging of non-small-cell lung cancer.Cancer Control. 2001; 8: 311-317PubMed Google Scholar, 7.White Jr., P. Ettinger D.S. Tissue is the issue: is endosonographic ultrasonography with or without fine-needle aspiration biopsy in the staging of non-small-cell lung cancer an advance?.Ann Intern Med. 1997; 127: 643-645Crossref PubMed Google Scholar, 8.Wallace M.B. Fritscher-Ravens A. Savides T.J. Endoscopic ultrasound for the staging of non-small-cell lung cancer.Endoscopy. 2003; 35: 606-610Crossref PubMed Scopus (30) Google Scholar, 9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar Prenzel et al.10.Prenzel K.L. Monig S.P. Sinning J.M. Baldus S.E. Brochhagen H.G. Schneider P.M. et al.Lymph node size and metastatic infiltration in non-small cell lung cancer.Chest. 2003; 123: 463-467Crossref PubMed Scopus (157) Google Scholar histologically evaluated 2891 lymph nodes from 256 NSCLC patients at the time of surgical resection of a primary tumor. Patients with locally advanced disease with clear-cut bulky mediastinal involvement were not surgical candidates and were excluded from the study. The average size of a malignant lymph node was 10.7±4.7 mm; 44% of these were 9 mm or less in diameter, and 12% of patients with lymph node metastasis had no lymph nodes greater than 10 mm in diameter. Studies have shown positron emission tomography (PET) scanning to have a superior sensitivity (67%-100%), specificity (81%-100%), and accuracy (79%-100%) over CT in defining lymph nodal metastasis.6.Lloyd C. Silvestri G.A. Mediastinal staging of non-small-cell lung cancer.Cancer Control. 2001; 8: 311-317PubMed Google Scholar, 9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar A major limitation of positron emission tomography (PET) is that fluorodeoxyglucose (FDG) uptake can be localized to the hilar or mediastinal areas but does not allow differentiation of left-sided from right-sided activity.9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar, 11.Gupta N. Graeber G.M. Bishop H.A. Comparative efficacy of positron emission tomography with fluorodeoxyglucose in evaluation of small (<1 cm), intermediate (1 to 3 cm) and large (>3 cm) lymph node lesion.Chest. 2000; 117: 773-778Crossref PubMed Scopus (214) Google Scholar This defect in 18-FDG PET detection of mediastinal involvement in NSCLC may be obviated with the application of CT-PET combination technology, which allows finer localization of the radiopharmaceutical uptake.EUS can easily detect lymph nodes in the aortopulmonic region (level 5), subcarinal region (level 7), paraesophageal region (level 8), and in the posterior mediastinum. EUS-FNA allows for definitive diagnosis of metastasis. Various studies have reported sensitivities for detection of mediastinal nodal metastasis in NSCLC ranging from 87% to 100%, with a specificity of 81% to 100%.5.Silvestri G.A. Hoffman B.J. Bhutani M.S. Hawes R.H. Coppage L. Sanders-Cliette A. et al.Endoscopic ultrasound with fine needle aspiration in the diagnosis and staging of lung cancer.Ann Thorac Surg. 1996; 61: 1441-1446Abstract Full Text PDF PubMed Scopus (258) Google Scholar, 8.Wallace M.B. Fritscher-Ravens A. Savides T.J. Endoscopic ultrasound for the staging of non-small-cell lung cancer.Endoscopy. 2003; 35: 606-610Crossref PubMed Scopus (30) Google Scholar, 9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar, 12.Larsen S.S. Krasnik M. Vilmann P. Jacobsen G.K. Pedersen J.H. Faurschou P. et al.Endoscopic ultrasound guided biopsy of mediastinal lesions has a major impact on patient management.Thorax. 2002; 57: 98-103Crossref PubMed Scopus (191) Google Scholar, 13.Fritscher-Ravens A. Soehendra N. Schirrow L. Sriram P.V. Meyer A. Hauber H.P. et al.Role of transesophageal endosonography-guided fine-needle aspiration in the diagnosis of lung cancer.Chest. 2000; 117: 339-345Crossref PubMed Scopus (180) Google Scholar, 14.Wiersema M.J. Vazquez-Sequerios E. Wiersema L.M. Evaluation of mediastinal lymphadenopathy with endoscopic US-guided fine-needle aspiration biopsy.Radiology. 2001; 219: 252-257Crossref PubMed Scopus (167) Google Scholar, 15.Gress F.G. Savides T.J. Sandler A. Kesler K. Conces D. Cummings O. et al.Endoscopic ultrasonography, fine-needle aspiration biopsy guided by endoscopic ultrasonography, and computed tomography in the preoperative staging of non-small-cell lung cancer: a comparison study.Ann Intern Med. 1997; 127: 604-612Crossref PubMed Google Scholar Fritscher-Ravens et al.9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar demonstrated EUS-FNA to be superior to the combination of PET and CT. Although EUS-FNA allows for sampling of lymph nodes as small as 5 mm, most false-negative results occur with aspiration of lymph nodes less than 10 mm in diameter with 1- to 2-mm foci of metastatic disease.5.Silvestri G.A. Hoffman B.J. Bhutani M.S. Hawes R.H. Coppage L. Sanders-Cliette A. et al.Endoscopic ultrasound with fine needle aspiration in the diagnosis and staging of lung cancer.Ann Thorac Surg. 1996; 61: 1441-1446Abstract Full Text PDF PubMed Scopus (258) Google Scholar, 15.Gress F.G. Savides T.J. Sandler A. Kesler K. Conces D. Cummings O. et al.Endoscopic ultrasonography, fine-needle aspiration biopsy guided by endoscopic ultrasonography, and computed tomography in the preoperative staging of non-small-cell lung cancer: a comparison study.Ann Intern Med. 1997; 127: 604-612Crossref PubMed Google Scholar A negative EUS-FNA still requires further evaluation, which may occur with adjunctive molecular techniques or at the time of operative resection. An EUS diagnosis of metastatic disease has a profound affect on the clinical management of patients; up to 80% may avoid a surgical intervention, depending upon the population studied.16.Luke W.P. Pearson F.G. Todd T.R. Patterson G.A. Cooper J.D. Prospective evaluation of mediastinoscopy for assessment of carcinoma of the lung.J Thorac Cardiovasc Surg. 1986; 91: 53-56PubMed Google ScholarTransbronchial needle aspiration (TBNA) can access paratracheal lymph nodes (level 2R, 4L, and 4R) that are not accessible by EUS-FNA. Transbronchial needle aspiration also may sample subcarinal lymph nodes (level 7). The overall sensitivity and specificity of TBNA has been reported to be 76% and 96%, respectively.17.Toloza E.M. Harpole L. Detterbeck F. McCrory D.C. Invasive staging of non-small cell lung cancer: a review of the current evidence.Chest. 2003; 123: 157S-166SCrossref PubMed Google Scholar The contribution of endobronchial US in increasing the yield of TBNA is undergoing evaluation at a number of centers. Until recently, mediastinoscopy has been the “reference standard” for sampling mediastinal lymph nodes. The sensitivity has been reported to be in the range of 70% to 95%.6.Lloyd C. Silvestri G.A. Mediastinal staging of non-small-cell lung cancer.Cancer Control. 2001; 8: 311-317PubMed Google Scholar, 7.White Jr., P. Ettinger D.S. Tissue is the issue: is endosonographic ultrasonography with or without fine-needle aspiration biopsy in the staging of non-small-cell lung cancer an advance?.Ann Intern Med. 1997; 127: 643-645Crossref PubMed Google Scholar, 16.Luke W.P. Pearson F.G. Todd T.R. Patterson G.A. Cooper J.D. Prospective evaluation of mediastinoscopy for assessment of carcinoma of the lung.J Thorac Cardiovasc Surg. 1986; 91: 53-56PubMed Google Scholar Mediastinoscopy can access lymph nodes in the anterior mediastinum that are not accessible by EUS-FNA, as well as lymph nodes in the subcarinal area and aortopulmonic window. A decision analysis model of patients with lung cancer and mediastinal lymphadenopathy at levels 5, 6, and 7 (accessible to both EUS-FNA and mediastinoscopy) found on CT revealed EUS-FNA to be cost-effective compared with mediastinoscopy.18.Aabakken L. Silvestri G.A. Hawes R. Reed C.E. Marsi V. Hoffman B. Cost-efficacy of endoscopic ultrasonography with fine-needle aspiration vs. mediastinotomy in patients with lung cancer and suspected mediastinal adenopathy.Endoscopy. 1999; 31: 707-711Crossref PubMed Scopus (105) Google ScholarSome issues with the study of Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar deserve further attention. They define T4 NSCLC as involving direct mediastinal invasion (i.e., invasion of central mediastinal structures, such as the great vessels and esophagus) and/or the presence of malignant pleural effusion; the T4 classification also includes tumors invading the trachea, carina, and vertebral body, as well as those tumors with intralobar satellite lesions. EUS can detect posterior involvement of the trachea and carina and, potentially, involvement of the vertebral body but is limited in evaluation of the right-sided and anterior portions of the central airways and cannot detect satellite lesions because of the tissue-air interface of the lung. The criteria Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar used, whereby, direct visualization of the tumor indicates invasion is not intuitive; atelectasis, chronic airway obstruction, and other reasons for paucity of air in the lung parenchyma could result in the same finding. Involvement of the great vessels is not defined with respect to which EUS criteria were used.The frequency of T4 disease, absent documented N2 or N3 lymph nodal disease is low, probably less than 25% of all T4 tumors. Nonetheless, some patients with T4 N0 M0 NSCLC may still be candidates for definitive surgical resection, particularly if their T staging was on the basis of carinal involvement. In a series from Japan, patients with T4 NSCLC who were able to undergo a complete resection at the time of surgery, had a 5-year survival rate as high as 22%.19.Tsuchiya R. Asamura H. Kondo H. Goya T. Naruke T. Extended resection of the left atrium, great vessels, or both for lung cancer.Ann Thorac Surg. 1994; 57: 960-965Abstract Full Text PDF PubMed Scopus (139) Google ScholarPatients with pleural effusions, discussed by Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar as an exclusion criterion to EUS staging, were studied with EUS in this paper. Why they were not excluded was not explained, nor was a definition offered for “large pleural effusion.” Simple tumor staging of patients with potentially malignant pleural effusions can be accomplished easily with diagnostic thoracentesis. Recalculation of the positive predictive value, leaving the two patients with pleural effusion out, finds this decreased to 62.5%.The study of Varadarajulu et al.,3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar along with others cited by these investigators, indicates that there are situations and referral biases that may render the institution-specific utility of EUS of lesser value for T staging than that implied in earlier studies. Although it may be acceptable to view the study of Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar as a negative result that may be disproven later, this is unlikely. By their honesty, these investigators have performed a service for gastroenterologists and clinicians and advocates for our patients. We disagree with their conclusion that EUS staging of the primary tumor may be accomplished with the criteria that they recommend in their discussion, because further prospective validation of these criteria and their accuracy is needed.Ultimately, the combination of TBFNA with endobronchial US and EUS-FNA may provide the most reasonable and complete approach to the non-surgical staging and diagnosis of NSCLC, while avoiding invasive procedures, including mediastinoscopy. At this point in time, while the use of routine EUS T staging of primary tumors in the setting of NSCLC is best avoided, it should not be forgotten that the body of evidence speaks loudly that the use of EUS-FNA for lymph nodal staging is a valuable technique in the correct clinical setting when performed by gastroenterologists who are familiar with the clinical disease. EUS-FNA for the primary diagnosis of lung cancer and for the lymph nodal staging of NSCLC has already been demonstrated to be accurate, cost-effective, and safe. A thorough understanding of lung cancer staging principles is needed, including a facile working knowledge of the lymph nodal staging of NSCLC, which is not similar to the staging of primary GI malignancies; the skills may not be easily transferable. Because most of the published information in the United States stems from gastroenterologists who share a similar training background in EUS, a gastroenterologist who seeks to perform lung cancer staging may need to seek additional knowledge and training from colleagues and nationally recognized experts. A final caveat is that gastroenterologists performing EUS and EUS-FNA for lung cancer should have a working relationship with the pulmonologists and the thoracic surgeons, so that the information sought and subsequently conveyed is clinically meaningful and accurate, and not provided solely as a technical procedure. The use of EUS in the staging and diagnosis of lung carcinoma has evolved over the past decade but not as rapidly nor as widely, compared with its GI applications. Since the first descriptions of EUS-guided FNA (EUS-FNA) in the setting of lung carcinoma,1.Wiersema M.J. Kochman M.L. Chak A. Cramer H.M. Kessler K. Real-time endoscopic ultrasound guided fine needle aspiration of a mediastinal lymph node.Gastrointest Endosc. 1993; 39: 429-431Abstract Full Text PDF PubMed Scopus (75) Google Scholar, 2.Wiersema M.J. Kochman M.L. Cramer H.M. Wiersema L.M. Preoperative staging of non-small cell lung cancer: transesophageal US-guided fine-needle aspiration biopsy of mediastinal lymph nodes.Radiology. 1994; 190: 239-242Crossref PubMed Scopus (52) Google Scholar a number of studies have been published describing the utility of EUS-FNA in the lymph nodal staging of non-small-cell lung carcinoma (NSCLC) and have found EUS to be cost-effective in the pre-resection staging evaluation for NSCLC. A few articles have noted primary tumor staging results, but none of these described a systematic approach with verification of the accuracy of the staging. The publication in this issue of Gastrointestinal Endoscopy by Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar is the first attempt to verify the accuracy of primary tumor staging of NSCLC when performed with EUS. In this retrospective study, performed by querying a database of 308 patients that contained pertinent information from the staging procedures that were performed at the Medical University of South Carolina for lymph nodal staging and tissue diagnosis, these investigators demonstrated that the accuracy of T4 staging was below that which traditionally would be expected for a staging modality. Because surgical resection remains the cornerstone of curative therapy for NSCLC, even with locally advanced disease, it is bothersome that inaccurate T staging by EUS may prohibit patients from undergoing a potentially curative surgical procedure. Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar provide data from a subset of their registry, with data gathered from 175 patients with surgical staging, 10 of whom were found to have T4 disease by EUS staging, and one in that cohort who was not staged as T4 but subsequently was demonstrated to have T4 disease at the time of surgery. For T4 lesions, these data yield a positive predictive value of 70%, a sensitivity of 87.5%, and a specificity of 98%. The specific patient population studied and the selection process used to refer the patients for the EUS staging procedure may significantly affect these numbers. Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar do not have surgical confirmation for the remaining 133 patients, of whom 5 were staged as having T4 disease by EUS; of these, 3 were staged by CT as having T3 cancer, and two were staged by CT as having T4 cancer. Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar referenced the old TNM staging and did not use the current staging schema,4.Greene F.L. Page D.L. Fleming I.D. Fritz A. Balch C.M. Haller D.G. et al.AJCC cancer staging manual.6th ed. Springer, New York2002Crossref Google Scholar but there is no difference in the T and N staging or the stage grouping of these two schemas. It is important to recognize and understand that bulky N2 or N3 lymph nodal disease, T4 primary tumors (other than intralobar satellite lesions), and distant metastatic disease generally preclude surgical cure, even in the setting of multimodality therapy. In addition, it is necessary to examine the clinical context in which the staging of lung carcinoma is occurring. At our institution, the majority of patients with suspected or proven NSCLC referred for mediastinal evaluation with EUS were felt to be poor surgical candidates or were highly suspected to harbor a postoperative recurrence. Why this institutional selection bias occurs is an important issue and pertinent to the options for the cost-effective evaluation and proper staging of patients with NSCLC. We work in an environment that has an active therapeutic endoscopy program, a busy interventional bronchoscopy service, and thoracic surgeons who are well versed and skilled at minimally invasive diagnostic and therapeutic procedures, which allows for a multidisciplinary approach. The staging of lung carcinoma in the pre-operative setting may be accomplished in any number of ways, some complementary and some competitive, but all share the ultimate goal of selecting those patients who will either clearly benefit from a surgical intervention and those who will clearly not benefit from intervention. A brief review of the available staging modalities is helpful in placing the study by Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar in the correct clinical context. The most widely disseminated modality to detect and stage a patient with lung cancer is helical CT. CT is invaluable for detecting distant metastasis, but the evaluation of regional lymph node metastasis is limited. A CT criterion for metastatic disease includes lymph node size of 1 cm or greater, which clearly limits its sensitivity. The range of sensitivity of CT in detecting mediastinal lymph nodes ranges from 43% to 89%, with a specificity of 32% to 94%.5.Silvestri G.A. Hoffman B.J. Bhutani M.S. Hawes R.H. Coppage L. Sanders-Cliette A. et al.Endoscopic ultrasound with fine needle aspiration in the diagnosis and staging of lung cancer.Ann Thorac Surg. 1996; 61: 1441-1446Abstract Full Text PDF PubMed Scopus (258) Google Scholar, 6.Lloyd C. Silvestri G.A. Mediastinal staging of non-small-cell lung cancer.Cancer Control. 2001; 8: 311-317PubMed Google Scholar, 7.White Jr., P. Ettinger D.S. Tissue is the issue: is endosonographic ultrasonography with or without fine-needle aspiration biopsy in the staging of non-small-cell lung cancer an advance?.Ann Intern Med. 1997; 127: 643-645Crossref PubMed Google Scholar, 8.Wallace M.B. Fritscher-Ravens A. Savides T.J. Endoscopic ultrasound for the staging of non-small-cell lung cancer.Endoscopy. 2003; 35: 606-610Crossref PubMed Scopus (30) Google Scholar, 9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar Prenzel et al.10.Prenzel K.L. Monig S.P. Sinning J.M. Baldus S.E. Brochhagen H.G. Schneider P.M. et al.Lymph node size and metastatic infiltration in non-small cell lung cancer.Chest. 2003; 123: 463-467Crossref PubMed Scopus (157) Google Scholar histologically evaluated 2891 lymph nodes from 256 NSCLC patients at the time of surgical resection of a primary tumor. Patients with locally advanced disease with clear-cut bulky mediastinal involvement were not surgical candidates and were excluded from the study. The average size of a malignant lymph node was 10.7±4.7 mm; 44% of these were 9 mm or less in diameter, and 12% of patients with lymph node metastasis had no lymph nodes greater than 10 mm in diameter. Studies have shown positron emission tomography (PET) scanning to have a superior sensitivity (67%-100%), specificity (81%-100%), and accuracy (79%-100%) over CT in defining lymph nodal metastasis.6.Lloyd C. Silvestri G.A. Mediastinal staging of non-small-cell lung cancer.Cancer Control. 2001; 8: 311-317PubMed Google Scholar, 9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar A major limitation of positron emission tomography (PET) is that fluorodeoxyglucose (FDG) uptake can be localized to the hilar or mediastinal areas but does not allow differentiation of left-sided from right-sided activity.9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar, 11.Gupta N. Graeber G.M. Bishop H.A. Comparative efficacy of positron emission tomography with fluorodeoxyglucose in evaluation of small (<1 cm), intermediate (1 to 3 cm) and large (>3 cm) lymph node lesion.Chest. 2000; 117: 773-778Crossref PubMed Scopus (214) Google Scholar This defect in 18-FDG PET detection of mediastinal involvement in NSCLC may be obviated with the application of CT-PET combination technology, which allows finer localization of the radiopharmaceutical uptake. EUS can easily detect lymph nodes in the aortopulmonic region (level 5), subcarinal region (level 7), paraesophageal region (level 8), and in the posterior mediastinum. EUS-FNA allows for definitive diagnosis of metastasis. Various studies have reported sensitivities for detection of mediastinal nodal metastasis in NSCLC ranging from 87% to 100%, with a specificity of 81% to 100%.5.Silvestri G.A. Hoffman B.J. Bhutani M.S. Hawes R.H. Coppage L. Sanders-Cliette A. et al.Endoscopic ultrasound with fine needle aspiration in the diagnosis and staging of lung cancer.Ann Thorac Surg. 1996; 61: 1441-1446Abstract Full Text PDF PubMed Scopus (258) Google Scholar, 8.Wallace M.B. Fritscher-Ravens A. Savides T.J. Endoscopic ultrasound for the staging of non-small-cell lung cancer.Endoscopy. 2003; 35: 606-610Crossref PubMed Scopus (30) Google Scholar, 9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar, 12.Larsen S.S. Krasnik M. Vilmann P. Jacobsen G.K. Pedersen J.H. Faurschou P. et al.Endoscopic ultrasound guided biopsy of mediastinal lesions has a major impact on patient management.Thorax. 2002; 57: 98-103Crossref PubMed Scopus (191) Google Scholar, 13.Fritscher-Ravens A. Soehendra N. Schirrow L. Sriram P.V. Meyer A. Hauber H.P. et al.Role of transesophageal endosonography-guided fine-needle aspiration in the diagnosis of lung cancer.Chest. 2000; 117: 339-345Crossref PubMed Scopus (180) Google Scholar, 14.Wiersema M.J. Vazquez-Sequerios E. Wiersema L.M. Evaluation of mediastinal lymphadenopathy with endoscopic US-guided fine-needle aspiration biopsy.Radiology. 2001; 219: 252-257Crossref PubMed Scopus (167) Google Scholar, 15.Gress F.G. Savides T.J. Sandler A. Kesler K. Conces D. Cummings O. et al.Endoscopic ultrasonography, fine-needle aspiration biopsy guided by endoscopic ultrasonography, and computed tomography in the preoperative staging of non-small-cell lung cancer: a comparison study.Ann Intern Med. 1997; 127: 604-612Crossref PubMed Google Scholar Fritscher-Ravens et al.9.Fritscher-Ravens A. Bohuslavizki K.H. Brandt L. Bobrowski C. Lund C. Knofel W.T. et al.Mediastinal lymph node involvement in potentially resectable lung cancer: comparison of CT, positron emission tomography, and endosonographic ultrasonography with and without fine-needle aspiration.Chest. 2003; 123: 442-451Crossref PubMed Scopus (194) Google Scholar demonstrated EUS-FNA to be superior to the combination of PET and CT. Although EUS-FNA allows for sampling of lymph nodes as small as 5 mm, most false-negative results occur with aspiration of lymph nodes less than 10 mm in diameter with 1- to 2-mm foci of metastatic disease.5.Silvestri G.A. Hoffman B.J. Bhutani M.S. Hawes R.H. Coppage L. Sanders-Cliette A. et al.Endoscopic ultrasound with fine needle aspiration in the diagnosis and staging of lung cancer.Ann Thorac Surg. 1996; 61: 1441-1446Abstract Full Text PDF PubMed Scopus (258) Google Scholar, 15.Gress F.G. Savides T.J. Sandler A. Kesler K. Conces D. Cummings O. et al.Endoscopic ultrasonography, fine-needle aspiration biopsy guided by endoscopic ultrasonography, and computed tomography in the preoperative staging of non-small-cell lung cancer: a comparison study.Ann Intern Med. 1997; 127: 604-612Crossref PubMed Google Scholar A negative EUS-FNA still requires further evaluation, which may occur with adjunctive molecular techniques or at the time of operative resection. An EUS diagnosis of metastatic disease has a profound affect on the clinical management of patients; up to 80% may avoid a surgical intervention, depending upon the population studied.16.Luke W.P. Pearson F.G. Todd T.R. Patterson G.A. Cooper J.D. Prospective evaluation of mediastinoscopy for assessment of carcinoma of the lung.J Thorac Cardiovasc Surg. 1986; 91: 53-56PubMed Google Scholar Transbronchial needle aspiration (TBNA) can access paratracheal lymph nodes (level 2R, 4L, and 4R) that are not accessible by EUS-FNA. Transbronchial needle aspiration also may sample subcarinal lymph nodes (level 7). The overall sensitivity and specificity of TBNA has been reported to be 76% and 96%, respectively.17.Toloza E.M. Harpole L. Detterbeck F. McCrory D.C. Invasive staging of non-small cell lung cancer: a review of the current evidence.Chest. 2003; 123: 157S-166SCrossref PubMed Google Scholar The contribution of endobronchial US in increasing the yield of TBNA is undergoing evaluation at a number of centers. Until recently, mediastinoscopy has been the “reference standard” for sampling mediastinal lymph nodes. The sensitivity has been reported to be in the range of 70% to 95%.6.Lloyd C. Silvestri G.A. Mediastinal staging of non-small-cell lung cancer.Cancer Control. 2001; 8: 311-317PubMed Google Scholar, 7.White Jr., P. Ettinger D.S. Tissue is the issue: is endosonographic ultrasonography with or without fine-needle aspiration biopsy in the staging of non-small-cell lung cancer an advance?.Ann Intern Med. 1997; 127: 643-645Crossref PubMed Google Scholar, 16.Luke W.P. Pearson F.G. Todd T.R. Patterson G.A. Cooper J.D. Prospective evaluation of mediastinoscopy for assessment of carcinoma of the lung.J Thorac Cardiovasc Surg. 1986; 91: 53-56PubMed Google Scholar Mediastinoscopy can access lymph nodes in the anterior mediastinum that are not accessible by EUS-FNA, as well as lymph nodes in the subcarinal area and aortopulmonic window. A decision analysis model of patients with lung cancer and mediastinal lymphadenopathy at levels 5, 6, and 7 (accessible to both EUS-FNA and mediastinoscopy) found on CT revealed EUS-FNA to be cost-effective compared with mediastinoscopy.18.Aabakken L. Silvestri G.A. Hawes R. Reed C.E. Marsi V. Hoffman B. Cost-efficacy of endoscopic ultrasonography with fine-needle aspiration vs. mediastinotomy in patients with lung cancer and suspected mediastinal adenopathy.Endoscopy. 1999; 31: 707-711Crossref PubMed Scopus (105) Google Scholar Some issues with the study of Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar deserve further attention. They define T4 NSCLC as involving direct mediastinal invasion (i.e., invasion of central mediastinal structures, such as the great vessels and esophagus) and/or the presence of malignant pleural effusion; the T4 classification also includes tumors invading the trachea, carina, and vertebral body, as well as those tumors with intralobar satellite lesions. EUS can detect posterior involvement of the trachea and carina and, potentially, involvement of the vertebral body but is limited in evaluation of the right-sided and anterior portions of the central airways and cannot detect satellite lesions because of the tissue-air interface of the lung. The criteria Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar used, whereby, direct visualization of the tumor indicates invasion is not intuitive; atelectasis, chronic airway obstruction, and other reasons for paucity of air in the lung parenchyma could result in the same finding. Involvement of the great vessels is not defined with respect to which EUS criteria were used. The frequency of T4 disease, absent documented N2 or N3 lymph nodal disease is low, probably less than 25% of all T4 tumors. Nonetheless, some patients with T4 N0 M0 NSCLC may still be candidates for definitive surgical resection, particularly if their T staging was on the basis of carinal involvement. In a series from Japan, patients with T4 NSCLC who were able to undergo a complete resection at the time of surgery, had a 5-year survival rate as high as 22%.19.Tsuchiya R. Asamura H. Kondo H. Goya T. Naruke T. Extended resection of the left atrium, great vessels, or both for lung cancer.Ann Thorac Surg. 1994; 57: 960-965Abstract Full Text PDF PubMed Scopus (139) Google Scholar Patients with pleural effusions, discussed by Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar as an exclusion criterion to EUS staging, were studied with EUS in this paper. Why they were not excluded was not explained, nor was a definition offered for “large pleural effusion.” Simple tumor staging of patients with potentially malignant pleural effusions can be accomplished easily with diagnostic thoracentesis. Recalculation of the positive predictive value, leaving the two patients with pleural effusion out, finds this decreased to 62.5%. The study of Varadarajulu et al.,3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar along with others cited by these investigators, indicates that there are situations and referral biases that may render the institution-specific utility of EUS of lesser value for T staging than that implied in earlier studies. Although it may be acceptable to view the study of Varadarajulu et al.3.Varadarajulu S. Schmulewitz N. Wildi S.F. Roberts S. Ravenel J. Reed C.E. et al.Accuracy of EUS in T4 staging in lung cancer.Gastrointest Endosc. 2004; 59: 345-348Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar as a negative result that may be disproven later, this is unlikely. By their honesty, these investigators have performed a service for gastroenterologists and clinicians and advocates for our patients. We disagree with their conclusion that EUS staging of the primary tumor may be accomplished with the criteria that they recommend in their discussion, because further prospective validation of these criteria and their accuracy is needed. Ultimately, the combination of TBFNA with endobronchial US and EUS-FNA may provide the most reasonable and complete approach to the non-surgical staging and diagnosis of NSCLC, while avoiding invasive procedures, including mediastinoscopy. At this point in time, while the use of routine EUS T staging of primary tumors in the setting of NSCLC is best avoided, it should not be forgotten that the body of evidence speaks loudly that the use of EUS-FNA for lymph nodal staging is a valuable technique in the correct clinical setting when performed by gastroenterologists who are familiar with the clinical disease. EUS-FNA for the primary diagnosis of lung cancer and for the lymph nodal staging of NSCLC has already been demonstrated to be accurate, cost-effective, and safe. A thorough understanding of lung cancer staging principles is needed, including a facile working knowledge of the lymph nodal staging of NSCLC, which is not similar to the staging of primary GI malignancies; the skills may not be easily transferable. Because most of the published information in the United States stems from gastroenterologists who share a similar training background in EUS, a gastroenterologist who seeks to perform lung cancer staging may need to seek additional knowledge and training from colleagues and nationally recognized experts. A final caveat is that gastroenterologists performing EUS and EUS-FNA for lung cancer should have a working relationship with the pulmonologists and the thoracic surgeons, so that the information sought and subsequently conveyed is clinically meaningful and accurate, and not provided solely as a technical procedure.