Endoscopic Ultrasound–Guided Fine-Needle Aspiration in the Diagnosis and Staging of Lung Cancer and Its Impact on Surgical Staging
http://www.100md.com
《临床肿瘤学》
the Departments of Pulmonary Medicine, Cardio-thoracic Surgery and Pathology, Leiden University Medical Center, Leiden, the Netherlands
ABSTRACT
PURPOSE: The diagnosis and staging of lung cancer critically depends on surgical procedures. Endoscopic ultrasound (EUS) –guided fine-needle aspiration (FNA) is an accurate, safe, and minimally invasive technique for the analysis of mediastinal lymph nodes (LNs) and can additionally detect tumor invasion (T4) in patients with centrally located tumors. The goal of this study was to assess to what extent EUS-FNA could prevent surgical interventions.
PATIENTS AND METHODS: Two hundred forty two consecutive patients with suspected (n = 142) or proven (n = 100) lung cancer and enlarged (> 1 cm) mediastinal LNs at chest computed tomography were scheduled for mediastinoscopy/tomy (94%) or exploratory thoracotomy (6%). Before surgery, all patients underwent EUS-FNA. If EUS-FNA established LN metastases, tumor invasion, or small-cell lung cancer (SCLC), scheduled surgical interventions were cancelled. Surgical-pathologic verification occurred when EUS-FNA did not demonstrate advanced disease. Cancelled surgical interventions because of EUS findings was the primary end point.
RESULTS: EUS-FNA prevented 70% of scheduled surgical procedures because of the demonstration of LN metastases in non–small-cell lung cancer (52%), tumor invasion (T4) (4%), tumor invasion and LN metastases (5%), SCLC (8%), or benign diagnoses (1%). Sensitivity, specificity, and accuracy for EUS in mediastinal analysis were 91%, 100% and 93%, respectively. No complications were recorded.
CONCLUSION: EUS-FNA qualifies as the initial staging procedure of choice for patients with (suspected) lung cancer and enlarged mediastinal LNs. Implementation of EUS-FNA in staging algorithms for lung cancer might reduce the number of surgical staging procedures considerably.
INTRODUCTION
Diagnosis and staging of lung cancer critically determines the proper treatment modality for these patients. To date, accurate mediastinal staging depends on invasive surgical interventions, such as mediastinoscopy. Tissue verification of mediastinal lymph nodes (LNs) suspect for metastases is important as patients with regional LN metastases are preferably treated with multimodality treatment, whereas, surgery alone is the treatment of choice in patients without regional and distant metastases.1 Mediastinoscopy is the standard method for mediastinal LN staging1 but drawbacks are its invasiveness, requirement of general anesthesia, need for clinical admission, as well as the risk of complications (2%).2
Endoscopic ultrasound (EUS) –guided fine-needle aspiration (FNA) has a sensitivity of 88% and specificity of 91% for the analysis of enlarged LNs in patients with non—small-cell lung cancer (NSCLC).3-5 In addition to LN staging, one study observed a high accuracy of EUS for the assessment of mediastinal tumor invasion (T4).6 Importantly, EUS-FNA is safe, minimally invasive, and performed in an ambulatory setting. Recently, a few medium-sized studies reported that EUS prevented mediastinoscopies in patients both with7,8 and without9 enlarged mediastinal LNs.
Large prospective studies evaluating the impact of EUS-FNA on surgical staging, by assessing mediastinal metastases or tumor invasion, are lacking. Therefore, we prospectively performed EUS-FNA in patients with (suspected) lung cancer and determined its impact on scheduled surgical staging procedures.
PATIENTS AND METHODS
Study Design
Patients with suspected or proven NSCLC and enlarged (> 1 cm short axis) mediastinal LNs at computed tomography (CT) scan of the chest—regardless of their localization—who were scheduled for a surgical staging procedure (mediastinoscopy/-tomy, (exploratory) thoracotomy) were eligible for this study. Staging by EUS-FNA before the intended surgery occurred under the assumption that EUS could prevent 50% of these interventions. Consecutive patients underwent EUS-FNA after obtaining informed consent. Surgical-pathologic staging occurred in those patients in whom neither LN metastases nor tumor invasion were found at EUS. Prevented surgical interventions because of EUS findings were recorded.
Patients
Two hundred forty-two consecutive patients were included in this study. The mean age of the study population was 63 years (range, 39 to 78) and 69% were male. The intrapulmonary tumors were evenly distributed between the left lung (44%) and right lung (44%)—left upper lobe 23%, left lower lobe 20%, lingula 1%, right upper lobe 17%, right lower lobe 26%, middle lobe 1%—and 12% of patients only presented with a mediastinal mass. Enlarged mediastinal LNs, as detected by CT scan of the chest, were located in regions considered to be within reach of EUS-FNA in 86% of patients (either paratracheally on the left, the aortopulmonary window, subcarinally, or in the lower mediastinum). In 16% of patients, enlarged mediastinal LNs were only located paratracheally on the right—a region considered difficult or beyond reach of EUS-FNA. The mean size of the mediastinal LNs at chest CT was 24 mm (range, 13 to 77 mm; Table 1). All patients had undergone at least one bronchoscopy previously and in 6% of patients transbronchial needle aspiration (TBNA) had been performed. The ethical committee of the Leiden University Medical Center (LUMC) approved the use of EUS-FNA for the staging of lung cancer. Included in the present study are 19 patients with PET positive mediastinal lesions that have been reported previously.10
EUS-FNA
EUS-FNA examinations were performed at the department of Pulmonary Medicine of the LUMC by two experienced investigators (JTA/KFR). The LUMC is a university and tertiary care hospital and the pulmonary department is a referral site for the analysis of mediastinal lesions by EUS-FNA. A Pentax FG 34 UX echo-endoscope (Pentax GmbH, Hamburg, Germany) with a longitudinal convex ultrasound transducer with an adjustable ultrasonic frequency of 5, 7.5 or 10 MHz was used in combination with a Hitachi EUB 6500 ultrasound scanner (Hitachi Medical Systems Ltd, Reeuwyk, the Netherlands). EUS-FNA was performed on an outpatient basis under conscious sedation using midazolam (1-5 mg IV). Before EUS-FNA, a CT scan of the chest was available to the investigators. At EUS, we started with a thorough investigation of all mediastinal LNs that can be reached from the esophagus. Enlarged or otherwise suspect LNs, were numbered according to the American Joint Committee on Cancer staging system11 and subsequently biopsied. If present, N3 LN stations were biopsied first, when ipsilateral LNs were visualized we started biopsying the most suspicious one obviously using a new needle. Aspirates were obtained under EUS guidance from the esophagus with a 22-gauge needle using suction (GIP/MEDI-Globe, type Hancke/Vilmann). The aspirated material was examined on site to judge the sample for its adequacy. After the procedure, all slides were examined by an experienced cytopathologist (MV). It was recorded whether the primary tumor was visible from the esophagus and, if so, whether invasion in centrally-located vessels or mediastinum was present (T4). We defined mediastinal tumor invasion as a tumor spreading continuously from its primary location towards the mediastinum on the basis of echographic appearance. Secondly, since this procedure has the major advantage of being performed under mild sedation, patients were asked to take a deep breath. If the tumor appeared to be fixed to the mediastinum of mediastinal organs, a suspicion of invasion was given. Patients were observed for 2 hours after the procedure. They were instructed to contact the hospital in case of chest discomfort or complaints.
Data Analysis
The primary end point of the study was cancellation of surgical interventions because of EUS findings. For LN staging, the final diagnosis was either on the basis of the LN aspirates revealing malignancy or on surgical-pathologic examination in those patients who underwent subsequent mediastinoscopy or thoracotomy. Sensitivity, specificity, accuracy, and positive and negative predictive values of EUS-FNA in mediastinal staging were obtained, excluding those patients where no surgical verification of tumor negative EUS results was obtained. Tumor positive EUS findings were not surgically validated as it was judged unethical to do so. False-positive EUS aspirates were considered unlikely, as in our study no cases existed where the aspirated LN and the primary tumor were located immediately adjacent to each other, a situation that could potentially lead to false-positive results if the tumor itself, instead of the LN, was biopsied. Statistical analysis was performed with SPSS, version 12 (SPSS, SPSS Inc., Chicago, IL).
RESULTS
EUS-FNA
EUS-FNA demonstrated LN metastases of NSCLC in 52% of patients, tumor invasion (T4) in 4% of patients, both tumor invasion and LN metastases (T4N2) in 5% of patients, and small-cell lung cancer (SCLC) in 8% of patients. In 1% of patients, an alternative diagnosis was established (ie, sarcoidosis n = 1, tuberculosis n = 1, carcinoid tumor n = 1). On the basis of these EUS findings, surgical interventions were cancelled in 70% of patients (Table 2). Additionally, separate analyses were performed for the group with a known NSCLC (n = 100) and the group without a prior diagnosis (n = 142). In the latter group the final description was NSCLC (n = 119), SCLC (n = 20), and non- malignant lesions (n = 3). For these two groups, 65% (staging NSCLC) and 73% (diagnosing suspected lung cancer) of surgical procedures were prevented because of EUS findings, respectively.
In 3% of patients, LNs were identified by EUS at station 5 (aortopulmonary window) or station 6 (para-aortal), however no biopsy was performed because of interposition of the pulmonary artery or aorta. Reactive LN tissue was found in 27% of patients, and in 3% of patients, the aspirates were inconclusive or contained not representative material. In 75% of patients, one mediastinal LN station, in 22% of patients two different mediastinal LN stations, and in 0.5% of patients three different LN stations were biopsied. In 2.5% of patients, no EUS-guided LN biopsy occurred. The mean number of needle passes per patient was three (range, 0 to 8). The most often-aspirated LN station, in 62% of patients, was position 7 (subcarinal). Tumor invasion (T4) was observed in 22 patients (9%) of whom 13 patients also had LN metastases (Fig 1). All six patients with N3 metastases had enlarged LNs at CT but in another 10 patients with enlarged contralateral LNs (N3) at CT, LNs were reactive. Tumor invasion (T4) was assessed in 9% of patients on the basis of invasion in the left atrium or centrally-located large vessels (n = 7) or mediastinum (n = 15). No complications, such as pneumothorax, bleeding, or fever, occurred.
Surgery
Surgical verification of tumor negative EUS findings occurred in 55 of 73 patients (75%; Fig 1). EUS proved to be false-negative in 15 patients because of the presence of LN metastases of NSCLC that were located paratracheally on the left (n = 2) paratracheally on the right (n = 4) subcarinally (n = 2), and on the lower mediastinum (n = 1). Additionally, LNs in the aortopulmonary window that were suspect on EUS but could not be safely biopsied were metastatic as confirmed by mediastinotomy (n = 2). Two patients were diagnosed with a SCLC and another two turned out to have a lymphoma. In the other 40 patients, EUS findings of the absence of mediastinal metastases were confirmed by mediastinoscopy/tomy and thoracotomy.
No surgical verification of tumor negative EUS findings occurred in 18 patients (Fig 1) because of presumed —PET positive but not tissue proven—N2 metastases (n = 2), a tumor positive pleural biopsy (n = 1), pleuritis carcinomatosa (n = 2), distant metastases (n = 5) and clinical deterioration (n = 4) or death (n = 1) in the interval before the scheduled thoracotomy, respectively, the treating physician deciding to administer neoadjuvant chemotherapy (n = 2), and patient refusal for further treatment (n = 1). Additionally, no surgical verification occurred in the nine patients staged T4 by EUS. The CT scan of the chest was also suggestive for tumor invasion in five of these patients.
The prevalence of regional LN metastases in this study population was 71%. The sensitivity, specificity, positive and negative predictive values, and accuracy for EUS-FNA for mediastinal staging on the basis of 215 assessable patients (Fig 1), were 91%, 100%, 100%, 74%, and 93%, respectively.
DISCUSSION
EUS-FNA prevented 70% of scheduled surgical interventions in patients with (suspected) lung cancer because of the demonstration of mediastinal metastases or tumor invasion (69%) or by establishing an alternative diagnosis (1%).
To our knowledge, this is by far the largest prospective study of EUS-FNA in lung cancer, assessing its impact on surgical staging. The ability to diagnose lung cancer by establishing LN metastases has been reported previously.12-14 The large proportion of surgical interventions that could be avoided in our study confirms the results of Larsen et al7 who found in 84 patients with enlarged mediastinal LNs that 49% of planned thoracotomies/thoracoscopies and 68% of planned mediastinoscopies could be cancelled because of EUS results. In another recent study in a similar population of 59 patients who were considered candidates for mediastinoscopy, EUS-FNA detected malignant LNs in 39% of patients, and the intended surgery finally occurred in only 22% of patients.8
The sensitivity of EUS-FNA for mediastinal LN analysis of 91% in this study is comparable with the 88% as reported in a recent review.5 The fact that on site cytology was performed in our study may contribute to this high yield. The negative predictive value of EUS-FNA of 74% in our study was similar as expected (77%),5 and was a result of false-negative EUS findings in 15 patients. Five patients proved to have LN metastases in stations previously biopsied by EUS-FNA (locations 4L (2x), 7 (2x), and 8, respectively) and were obviously false-negative because of a sampling error. In two patients, enlarged LNs with ultrasound features suspect for malignancy were observed by EUS in the aortopulmonary window, but no aspirates were taken because of the interposition of the pulmonary artery. Subsequently, LN metastases were proven by mediastinotomy in both cases. In four patients, metastases were exclusively located in paratracheal LNs on the right, an area known to be difficult to investigate by EUS-FNA because of intervening air between the ultrasound transducer and the target area. Additionally, two patients were diagnosed with a SCLC, and two patients turned out to have a lymphoma. The latter diagnosis cannot be made accurately with 22-gauge needles, which were used in this study.
The present study confirms the observation by Varadarajulu et al6 that tumor invasion in the mediastinum or mediastinal organs (T4) can be assessed by EUS provided that the tumor is centrally-located and lies adjacent to the esophagus. Surgical verification of T4 findings by EUS, however, occurred in none of our patients and in only part of the patients described by Varadarajulu, so that the accuracy for EUS for this indication is not known at present. As T-overstaging by EUS has been described,6 extreme caution should be taken before staging a tumor as T4 as this could result in denying a patient a potentially curative thoracotomy.
The observations of our study are clinically relevant because surgical interventions are often performed in the diagnosis and staging of lung cancer. Up to 30% of bronchoscopies for suspected lung cancer are nondiagnostic,15 and if patients have enlarged or PET positive mediastinal LNs, mediastinoscopy is often performed. By targeting enlarged mediastinal LNs, both a tissue diagnosis and mediastinal staging can be accomplished if LN metastases are present and proven. In patients with proven NSCLC and enlarged or PET positive mediastinal LNs, tissue verification of mediastinal LNs is obligatory.1,16-18
Although mediastinoscopy, with a sensitivity of 81% and specificity of 100%,5 is regarded as the standard technique for mediastinal staging, there are limitations in its diagnostic reach. Additionally, a recent study demonstrated that only 40% of mediastinoscopies are performed adequately.19 Further limitations of mediastinoscopy are the need for general anesthesia, clinical admission, and thus, high costs as well as a reported morbidity in 2% of cases.2 In contrast, no complications have been described so far in LN staging by EUS-FNA.3-5 It should be noted that mediastinoscopy and EUS-FNA are complementary in their diagnostic reach with EUS providing access to those LNs located adjacent to the esophagus in the left paratracheal area, the aortopulmonary window, the subcarinal area, and the lower mediastinum.10 EUS has its limitations in visualization of LNs located in the paratracheal area on the right because of intervening air between the esophagus and the LNs in these areas. Therefore, in our experience, mediastinoscopy remains the technique of choice for the analysis of LNs located pre- and paratracheally on the right. We prefer EUS-FNA above mediastinoscopy, as the mediastinal stations to be examined can be reached by both techniques.
In a recent report, we suggested a novel staging strategy for lung cancer in which PET positive mediastinal lesions were analyzed by EUS-FNA. In 69% of patients, EUS-FNA demonstrated mediastinal involvement.10 These results were confirmed in another study that concluded additionally that EUS-FNA could replace more than half of the surgical staging procedures and reduce staging costs by 40%.20 Whether in patients with (suspected) lung cancer and enlarged mediastinal LNs the next diagnostic step should be PET or EUS-FNA is under debate. As the specificity of PET is only 78% in patients with enlarged LNs,21 one may argue to start with EUS-FNA, as EUS-FNA provides a tissue diagnosis in a single test. An alternative to mediastinoscopy in the future might be endobronchial ultrasound (EBUS) guided TBNA, a novel technique by which LN located adjacent to the tracheobronchial tree can be biopsied in a real-time ultrasound-guided fashion. In a recent study from Japan in which hilar and mediastinal LNs were analyzed by EBUS-TBNA, an accuracy of 97% was reported.22 To date, no study has compared EUS-FNA with EBUS-TBNA.
Several limitations apply to our study. Only patients with (suspected) NSCLC and enlarged mediastinal LNs were included in this study and our results, therefore, only apply for this group. LN metastases occur in approximately 20% of patients with a normal mediastinum at CT. In two recent studies, EUS-FNA demonstrated locally advanced disease in up to 22% of NSCLC patients with a normal mediastinum at CT.9,23 Furthermore, the prevalence of LN metastasis of 71% in this cohort was relatively high, which might provide a bias towards a higher rate of prevented surgical interventions. The high prevalence might be explained by the fact that in 27% of patients, mediastinal LNs were positive at FDG-PET. Finally, EUS investigators of this study were experienced (> 500 EUS-FNA procedures) and whether the observed findings are reproducible by less trained investigators has to be established in further studies.
In conclusion, in patients with (suspected) lung cancer and enlarged mediastinal LNs, EUS-FNA demonstrated advanced disease in the majority of patients by assessing LN metastases or tumor invasion. Implementation of EUS-FNA in the diagnosis and staging of lung cancer will reduce the number of surgical diagnostic and staging procedures considerably.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank W.J. Oudshoorn and A. van der Mey for logistical and computing assistance.
NOTES
Supported by a grant from the Leiden University Medical Center (LUMC). Technical support by Hitachi Ultrasound, Reeuwijk, the Netherlands.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Spira A, Ettinger DS: Multidisciplinary management of lung cancer. N Engl J Med 350(4):379-392, 2004
Hammoud ZT, Anderson RC, Meyers BF, et al: The current role of mediastinoscopy in the evaluation of thoracic disease. J Thorac Cardiovasc Surg 118(5):894-899, 1999
Wallace MB, Fritscher-Ravens A, Savides TJ: Endoscopic ultrasound for the staging of non-small-cell lung cancer. Endoscopy 35(7):606-610, 2003
Fritscher-Ravens A: Endoscopic ultrasound evaluation in the diagnosis and staging of lung cancer. Lung Cancer 41:259-267, 2003
Toloza EM, Harpole L, Detterbeck F, et al: Invasive staging of non-small cell lung cancer: A review of the current evidence. Chest 123:157S-166S, 2003
Varadarajulu S, Schmulewitz N, Wildi SF, et al: Accuracy of EUS in staging of T4 lung cancer. Gastrointest Endosc 59(3):345-348, 2004
Larsen SS, Krasnik M, Vilmann P, et al: Endoscopic ultrasound guided biopsy of mediastinal lesions has a major impact on patient management. Thorax 57:98-103, 2002
Savides TJ, Perricone A: Impact of EUS-guided FNA of enlarged mediastinal lymph nodes on subsequent thoracic surgery rates. Gastrointest Endosc 60(3):340-346, 2004
Leblanc JK, Devereaux BM, Imperiale TF, et al: Endoscopic ultrasound in non-small cell lung cancer and negative mediastinum on computed tomography. Am J Respir Crit Care Med 171:177-182, 2004
Annema JT, Hoekstra OS, Smit EF, et al: Towards a minimally invasive staging strategy in NSCLC: Analysis of PET positive mediastinal lesions by EUS-FNA. Lung Cancer 44:53-60, 2004
Mountain CF, Dresler CM: Regional lymph node classification for lung cancer staging. Chest 111(6):1718-1723, 1997
Fritscher-Ravens A, Sriram PV, Bobrowski C, et al: Mediastinal lymphadenopathy in patients with or without previous malignancy: EUS-FNA-based differential cytodiagnosis in 153 patients. Am J Gastroenterol 95:2278-2284, 2000
Wiersema MJ, Vazquez-Sequeiros E, Wiersema LM: Evaluation of mediastinal lymphadenopathy with endoscopic US-guided fine-needle aspiration biopsy. Radiology 219(1):252-257, 2001
Williams DB, Sahai AV, Aabakken L, et al: Endoscopic ultrasound guided fine needle aspiration biopsy: A large single centre experience. Gut 44:720-726, 1999
Mazzone P, Jain P, Arroliga AC, et al: Bronchoscopy and needle biopsy techniques for diagnosis and staging of lung cancer. Clin Chest Med 23(1):137-158, 2002
Pieterman RM, van Putten JW, Meuzelaar JJ, et al: Preoperative staging of non-small-cell lung cancer with positron-emission tomography. N Engl J Med 343:254-261, 2000
van Tinteren H, Hoekstra OS, Smit EF, et al: Effectiveness of positron emission tomography in the preoperative assessment of patients with suspected non-small-cell lung cancer: The PLUS multicentre randomised trial. Lancet 359:1388-1393, 2002
Detterbeck FC, DeCamp MM, Jr, Kohman LJ, et al: Lung cancer. Invasive staging: The guidelines. Chest 123(1 Suppl):167S-175S, 2003
Smulders SA, Smeenk FW, Janssen-Heijnen ML, et al: Surgical mediastinal staging in daily practice. Lung Cancer 47:243-251, 2005
Kramer H, van Putten JW, Post WJ, et al: Oesophageal endoscopic ultrasound with fine needle aspiration improves and simplifies the staging of lung cancer. Thorax 59(7):596-601, 2004
Gould MK, Kuschner WG, Rydzak CE, et al: Test performance of positron emission tomography and computed tomography for mediastinal staging in patients with non-small-cell lung cancer: A meta-analysis. Ann Intern Med 139:879-892, 2003
Yasufuku K, Chiyo M, Sekine Y, et al: Real-time endobronchial ultrasound-guided transbronchial needle aspiration of mediastinal and hilar lymph nodes. Chest 126(1):122-128, 2004
Wallace MB, Ravenel J, Block MI, et al: Endoscopic ultrasound in lung cancer patients with a normal mediastinum on computed tomography. Ann Thorac Surg 77(5):1763-1768, 2004(Jouke T. Annema, Michel I)
ABSTRACT
PURPOSE: The diagnosis and staging of lung cancer critically depends on surgical procedures. Endoscopic ultrasound (EUS) –guided fine-needle aspiration (FNA) is an accurate, safe, and minimally invasive technique for the analysis of mediastinal lymph nodes (LNs) and can additionally detect tumor invasion (T4) in patients with centrally located tumors. The goal of this study was to assess to what extent EUS-FNA could prevent surgical interventions.
PATIENTS AND METHODS: Two hundred forty two consecutive patients with suspected (n = 142) or proven (n = 100) lung cancer and enlarged (> 1 cm) mediastinal LNs at chest computed tomography were scheduled for mediastinoscopy/tomy (94%) or exploratory thoracotomy (6%). Before surgery, all patients underwent EUS-FNA. If EUS-FNA established LN metastases, tumor invasion, or small-cell lung cancer (SCLC), scheduled surgical interventions were cancelled. Surgical-pathologic verification occurred when EUS-FNA did not demonstrate advanced disease. Cancelled surgical interventions because of EUS findings was the primary end point.
RESULTS: EUS-FNA prevented 70% of scheduled surgical procedures because of the demonstration of LN metastases in non–small-cell lung cancer (52%), tumor invasion (T4) (4%), tumor invasion and LN metastases (5%), SCLC (8%), or benign diagnoses (1%). Sensitivity, specificity, and accuracy for EUS in mediastinal analysis were 91%, 100% and 93%, respectively. No complications were recorded.
CONCLUSION: EUS-FNA qualifies as the initial staging procedure of choice for patients with (suspected) lung cancer and enlarged mediastinal LNs. Implementation of EUS-FNA in staging algorithms for lung cancer might reduce the number of surgical staging procedures considerably.
INTRODUCTION
Diagnosis and staging of lung cancer critically determines the proper treatment modality for these patients. To date, accurate mediastinal staging depends on invasive surgical interventions, such as mediastinoscopy. Tissue verification of mediastinal lymph nodes (LNs) suspect for metastases is important as patients with regional LN metastases are preferably treated with multimodality treatment, whereas, surgery alone is the treatment of choice in patients without regional and distant metastases.1 Mediastinoscopy is the standard method for mediastinal LN staging1 but drawbacks are its invasiveness, requirement of general anesthesia, need for clinical admission, as well as the risk of complications (2%).2
Endoscopic ultrasound (EUS) –guided fine-needle aspiration (FNA) has a sensitivity of 88% and specificity of 91% for the analysis of enlarged LNs in patients with non—small-cell lung cancer (NSCLC).3-5 In addition to LN staging, one study observed a high accuracy of EUS for the assessment of mediastinal tumor invasion (T4).6 Importantly, EUS-FNA is safe, minimally invasive, and performed in an ambulatory setting. Recently, a few medium-sized studies reported that EUS prevented mediastinoscopies in patients both with7,8 and without9 enlarged mediastinal LNs.
Large prospective studies evaluating the impact of EUS-FNA on surgical staging, by assessing mediastinal metastases or tumor invasion, are lacking. Therefore, we prospectively performed EUS-FNA in patients with (suspected) lung cancer and determined its impact on scheduled surgical staging procedures.
PATIENTS AND METHODS
Study Design
Patients with suspected or proven NSCLC and enlarged (> 1 cm short axis) mediastinal LNs at computed tomography (CT) scan of the chest—regardless of their localization—who were scheduled for a surgical staging procedure (mediastinoscopy/-tomy, (exploratory) thoracotomy) were eligible for this study. Staging by EUS-FNA before the intended surgery occurred under the assumption that EUS could prevent 50% of these interventions. Consecutive patients underwent EUS-FNA after obtaining informed consent. Surgical-pathologic staging occurred in those patients in whom neither LN metastases nor tumor invasion were found at EUS. Prevented surgical interventions because of EUS findings were recorded.
Patients
Two hundred forty-two consecutive patients were included in this study. The mean age of the study population was 63 years (range, 39 to 78) and 69% were male. The intrapulmonary tumors were evenly distributed between the left lung (44%) and right lung (44%)—left upper lobe 23%, left lower lobe 20%, lingula 1%, right upper lobe 17%, right lower lobe 26%, middle lobe 1%—and 12% of patients only presented with a mediastinal mass. Enlarged mediastinal LNs, as detected by CT scan of the chest, were located in regions considered to be within reach of EUS-FNA in 86% of patients (either paratracheally on the left, the aortopulmonary window, subcarinally, or in the lower mediastinum). In 16% of patients, enlarged mediastinal LNs were only located paratracheally on the right—a region considered difficult or beyond reach of EUS-FNA. The mean size of the mediastinal LNs at chest CT was 24 mm (range, 13 to 77 mm; Table 1). All patients had undergone at least one bronchoscopy previously and in 6% of patients transbronchial needle aspiration (TBNA) had been performed. The ethical committee of the Leiden University Medical Center (LUMC) approved the use of EUS-FNA for the staging of lung cancer. Included in the present study are 19 patients with PET positive mediastinal lesions that have been reported previously.10
EUS-FNA
EUS-FNA examinations were performed at the department of Pulmonary Medicine of the LUMC by two experienced investigators (JTA/KFR). The LUMC is a university and tertiary care hospital and the pulmonary department is a referral site for the analysis of mediastinal lesions by EUS-FNA. A Pentax FG 34 UX echo-endoscope (Pentax GmbH, Hamburg, Germany) with a longitudinal convex ultrasound transducer with an adjustable ultrasonic frequency of 5, 7.5 or 10 MHz was used in combination with a Hitachi EUB 6500 ultrasound scanner (Hitachi Medical Systems Ltd, Reeuwyk, the Netherlands). EUS-FNA was performed on an outpatient basis under conscious sedation using midazolam (1-5 mg IV). Before EUS-FNA, a CT scan of the chest was available to the investigators. At EUS, we started with a thorough investigation of all mediastinal LNs that can be reached from the esophagus. Enlarged or otherwise suspect LNs, were numbered according to the American Joint Committee on Cancer staging system11 and subsequently biopsied. If present, N3 LN stations were biopsied first, when ipsilateral LNs were visualized we started biopsying the most suspicious one obviously using a new needle. Aspirates were obtained under EUS guidance from the esophagus with a 22-gauge needle using suction (GIP/MEDI-Globe, type Hancke/Vilmann). The aspirated material was examined on site to judge the sample for its adequacy. After the procedure, all slides were examined by an experienced cytopathologist (MV). It was recorded whether the primary tumor was visible from the esophagus and, if so, whether invasion in centrally-located vessels or mediastinum was present (T4). We defined mediastinal tumor invasion as a tumor spreading continuously from its primary location towards the mediastinum on the basis of echographic appearance. Secondly, since this procedure has the major advantage of being performed under mild sedation, patients were asked to take a deep breath. If the tumor appeared to be fixed to the mediastinum of mediastinal organs, a suspicion of invasion was given. Patients were observed for 2 hours after the procedure. They were instructed to contact the hospital in case of chest discomfort or complaints.
Data Analysis
The primary end point of the study was cancellation of surgical interventions because of EUS findings. For LN staging, the final diagnosis was either on the basis of the LN aspirates revealing malignancy or on surgical-pathologic examination in those patients who underwent subsequent mediastinoscopy or thoracotomy. Sensitivity, specificity, accuracy, and positive and negative predictive values of EUS-FNA in mediastinal staging were obtained, excluding those patients where no surgical verification of tumor negative EUS results was obtained. Tumor positive EUS findings were not surgically validated as it was judged unethical to do so. False-positive EUS aspirates were considered unlikely, as in our study no cases existed where the aspirated LN and the primary tumor were located immediately adjacent to each other, a situation that could potentially lead to false-positive results if the tumor itself, instead of the LN, was biopsied. Statistical analysis was performed with SPSS, version 12 (SPSS, SPSS Inc., Chicago, IL).
RESULTS
EUS-FNA
EUS-FNA demonstrated LN metastases of NSCLC in 52% of patients, tumor invasion (T4) in 4% of patients, both tumor invasion and LN metastases (T4N2) in 5% of patients, and small-cell lung cancer (SCLC) in 8% of patients. In 1% of patients, an alternative diagnosis was established (ie, sarcoidosis n = 1, tuberculosis n = 1, carcinoid tumor n = 1). On the basis of these EUS findings, surgical interventions were cancelled in 70% of patients (Table 2). Additionally, separate analyses were performed for the group with a known NSCLC (n = 100) and the group without a prior diagnosis (n = 142). In the latter group the final description was NSCLC (n = 119), SCLC (n = 20), and non- malignant lesions (n = 3). For these two groups, 65% (staging NSCLC) and 73% (diagnosing suspected lung cancer) of surgical procedures were prevented because of EUS findings, respectively.
In 3% of patients, LNs were identified by EUS at station 5 (aortopulmonary window) or station 6 (para-aortal), however no biopsy was performed because of interposition of the pulmonary artery or aorta. Reactive LN tissue was found in 27% of patients, and in 3% of patients, the aspirates were inconclusive or contained not representative material. In 75% of patients, one mediastinal LN station, in 22% of patients two different mediastinal LN stations, and in 0.5% of patients three different LN stations were biopsied. In 2.5% of patients, no EUS-guided LN biopsy occurred. The mean number of needle passes per patient was three (range, 0 to 8). The most often-aspirated LN station, in 62% of patients, was position 7 (subcarinal). Tumor invasion (T4) was observed in 22 patients (9%) of whom 13 patients also had LN metastases (Fig 1). All six patients with N3 metastases had enlarged LNs at CT but in another 10 patients with enlarged contralateral LNs (N3) at CT, LNs were reactive. Tumor invasion (T4) was assessed in 9% of patients on the basis of invasion in the left atrium or centrally-located large vessels (n = 7) or mediastinum (n = 15). No complications, such as pneumothorax, bleeding, or fever, occurred.
Surgery
Surgical verification of tumor negative EUS findings occurred in 55 of 73 patients (75%; Fig 1). EUS proved to be false-negative in 15 patients because of the presence of LN metastases of NSCLC that were located paratracheally on the left (n = 2) paratracheally on the right (n = 4) subcarinally (n = 2), and on the lower mediastinum (n = 1). Additionally, LNs in the aortopulmonary window that were suspect on EUS but could not be safely biopsied were metastatic as confirmed by mediastinotomy (n = 2). Two patients were diagnosed with a SCLC and another two turned out to have a lymphoma. In the other 40 patients, EUS findings of the absence of mediastinal metastases were confirmed by mediastinoscopy/tomy and thoracotomy.
No surgical verification of tumor negative EUS findings occurred in 18 patients (Fig 1) because of presumed —PET positive but not tissue proven—N2 metastases (n = 2), a tumor positive pleural biopsy (n = 1), pleuritis carcinomatosa (n = 2), distant metastases (n = 5) and clinical deterioration (n = 4) or death (n = 1) in the interval before the scheduled thoracotomy, respectively, the treating physician deciding to administer neoadjuvant chemotherapy (n = 2), and patient refusal for further treatment (n = 1). Additionally, no surgical verification occurred in the nine patients staged T4 by EUS. The CT scan of the chest was also suggestive for tumor invasion in five of these patients.
The prevalence of regional LN metastases in this study population was 71%. The sensitivity, specificity, positive and negative predictive values, and accuracy for EUS-FNA for mediastinal staging on the basis of 215 assessable patients (Fig 1), were 91%, 100%, 100%, 74%, and 93%, respectively.
DISCUSSION
EUS-FNA prevented 70% of scheduled surgical interventions in patients with (suspected) lung cancer because of the demonstration of mediastinal metastases or tumor invasion (69%) or by establishing an alternative diagnosis (1%).
To our knowledge, this is by far the largest prospective study of EUS-FNA in lung cancer, assessing its impact on surgical staging. The ability to diagnose lung cancer by establishing LN metastases has been reported previously.12-14 The large proportion of surgical interventions that could be avoided in our study confirms the results of Larsen et al7 who found in 84 patients with enlarged mediastinal LNs that 49% of planned thoracotomies/thoracoscopies and 68% of planned mediastinoscopies could be cancelled because of EUS results. In another recent study in a similar population of 59 patients who were considered candidates for mediastinoscopy, EUS-FNA detected malignant LNs in 39% of patients, and the intended surgery finally occurred in only 22% of patients.8
The sensitivity of EUS-FNA for mediastinal LN analysis of 91% in this study is comparable with the 88% as reported in a recent review.5 The fact that on site cytology was performed in our study may contribute to this high yield. The negative predictive value of EUS-FNA of 74% in our study was similar as expected (77%),5 and was a result of false-negative EUS findings in 15 patients. Five patients proved to have LN metastases in stations previously biopsied by EUS-FNA (locations 4L (2x), 7 (2x), and 8, respectively) and were obviously false-negative because of a sampling error. In two patients, enlarged LNs with ultrasound features suspect for malignancy were observed by EUS in the aortopulmonary window, but no aspirates were taken because of the interposition of the pulmonary artery. Subsequently, LN metastases were proven by mediastinotomy in both cases. In four patients, metastases were exclusively located in paratracheal LNs on the right, an area known to be difficult to investigate by EUS-FNA because of intervening air between the ultrasound transducer and the target area. Additionally, two patients were diagnosed with a SCLC, and two patients turned out to have a lymphoma. The latter diagnosis cannot be made accurately with 22-gauge needles, which were used in this study.
The present study confirms the observation by Varadarajulu et al6 that tumor invasion in the mediastinum or mediastinal organs (T4) can be assessed by EUS provided that the tumor is centrally-located and lies adjacent to the esophagus. Surgical verification of T4 findings by EUS, however, occurred in none of our patients and in only part of the patients described by Varadarajulu, so that the accuracy for EUS for this indication is not known at present. As T-overstaging by EUS has been described,6 extreme caution should be taken before staging a tumor as T4 as this could result in denying a patient a potentially curative thoracotomy.
The observations of our study are clinically relevant because surgical interventions are often performed in the diagnosis and staging of lung cancer. Up to 30% of bronchoscopies for suspected lung cancer are nondiagnostic,15 and if patients have enlarged or PET positive mediastinal LNs, mediastinoscopy is often performed. By targeting enlarged mediastinal LNs, both a tissue diagnosis and mediastinal staging can be accomplished if LN metastases are present and proven. In patients with proven NSCLC and enlarged or PET positive mediastinal LNs, tissue verification of mediastinal LNs is obligatory.1,16-18
Although mediastinoscopy, with a sensitivity of 81% and specificity of 100%,5 is regarded as the standard technique for mediastinal staging, there are limitations in its diagnostic reach. Additionally, a recent study demonstrated that only 40% of mediastinoscopies are performed adequately.19 Further limitations of mediastinoscopy are the need for general anesthesia, clinical admission, and thus, high costs as well as a reported morbidity in 2% of cases.2 In contrast, no complications have been described so far in LN staging by EUS-FNA.3-5 It should be noted that mediastinoscopy and EUS-FNA are complementary in their diagnostic reach with EUS providing access to those LNs located adjacent to the esophagus in the left paratracheal area, the aortopulmonary window, the subcarinal area, and the lower mediastinum.10 EUS has its limitations in visualization of LNs located in the paratracheal area on the right because of intervening air between the esophagus and the LNs in these areas. Therefore, in our experience, mediastinoscopy remains the technique of choice for the analysis of LNs located pre- and paratracheally on the right. We prefer EUS-FNA above mediastinoscopy, as the mediastinal stations to be examined can be reached by both techniques.
In a recent report, we suggested a novel staging strategy for lung cancer in which PET positive mediastinal lesions were analyzed by EUS-FNA. In 69% of patients, EUS-FNA demonstrated mediastinal involvement.10 These results were confirmed in another study that concluded additionally that EUS-FNA could replace more than half of the surgical staging procedures and reduce staging costs by 40%.20 Whether in patients with (suspected) lung cancer and enlarged mediastinal LNs the next diagnostic step should be PET or EUS-FNA is under debate. As the specificity of PET is only 78% in patients with enlarged LNs,21 one may argue to start with EUS-FNA, as EUS-FNA provides a tissue diagnosis in a single test. An alternative to mediastinoscopy in the future might be endobronchial ultrasound (EBUS) guided TBNA, a novel technique by which LN located adjacent to the tracheobronchial tree can be biopsied in a real-time ultrasound-guided fashion. In a recent study from Japan in which hilar and mediastinal LNs were analyzed by EBUS-TBNA, an accuracy of 97% was reported.22 To date, no study has compared EUS-FNA with EBUS-TBNA.
Several limitations apply to our study. Only patients with (suspected) NSCLC and enlarged mediastinal LNs were included in this study and our results, therefore, only apply for this group. LN metastases occur in approximately 20% of patients with a normal mediastinum at CT. In two recent studies, EUS-FNA demonstrated locally advanced disease in up to 22% of NSCLC patients with a normal mediastinum at CT.9,23 Furthermore, the prevalence of LN metastasis of 71% in this cohort was relatively high, which might provide a bias towards a higher rate of prevented surgical interventions. The high prevalence might be explained by the fact that in 27% of patients, mediastinal LNs were positive at FDG-PET. Finally, EUS investigators of this study were experienced (> 500 EUS-FNA procedures) and whether the observed findings are reproducible by less trained investigators has to be established in further studies.
In conclusion, in patients with (suspected) lung cancer and enlarged mediastinal LNs, EUS-FNA demonstrated advanced disease in the majority of patients by assessing LN metastases or tumor invasion. Implementation of EUS-FNA in the diagnosis and staging of lung cancer will reduce the number of surgical diagnostic and staging procedures considerably.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgment
We thank W.J. Oudshoorn and A. van der Mey for logistical and computing assistance.
NOTES
Supported by a grant from the Leiden University Medical Center (LUMC). Technical support by Hitachi Ultrasound, Reeuwijk, the Netherlands.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Spira A, Ettinger DS: Multidisciplinary management of lung cancer. N Engl J Med 350(4):379-392, 2004
Hammoud ZT, Anderson RC, Meyers BF, et al: The current role of mediastinoscopy in the evaluation of thoracic disease. J Thorac Cardiovasc Surg 118(5):894-899, 1999
Wallace MB, Fritscher-Ravens A, Savides TJ: Endoscopic ultrasound for the staging of non-small-cell lung cancer. Endoscopy 35(7):606-610, 2003
Fritscher-Ravens A: Endoscopic ultrasound evaluation in the diagnosis and staging of lung cancer. Lung Cancer 41:259-267, 2003
Toloza EM, Harpole L, Detterbeck F, et al: Invasive staging of non-small cell lung cancer: A review of the current evidence. Chest 123:157S-166S, 2003
Varadarajulu S, Schmulewitz N, Wildi SF, et al: Accuracy of EUS in staging of T4 lung cancer. Gastrointest Endosc 59(3):345-348, 2004
Larsen SS, Krasnik M, Vilmann P, et al: Endoscopic ultrasound guided biopsy of mediastinal lesions has a major impact on patient management. Thorax 57:98-103, 2002
Savides TJ, Perricone A: Impact of EUS-guided FNA of enlarged mediastinal lymph nodes on subsequent thoracic surgery rates. Gastrointest Endosc 60(3):340-346, 2004
Leblanc JK, Devereaux BM, Imperiale TF, et al: Endoscopic ultrasound in non-small cell lung cancer and negative mediastinum on computed tomography. Am J Respir Crit Care Med 171:177-182, 2004
Annema JT, Hoekstra OS, Smit EF, et al: Towards a minimally invasive staging strategy in NSCLC: Analysis of PET positive mediastinal lesions by EUS-FNA. Lung Cancer 44:53-60, 2004
Mountain CF, Dresler CM: Regional lymph node classification for lung cancer staging. Chest 111(6):1718-1723, 1997
Fritscher-Ravens A, Sriram PV, Bobrowski C, et al: Mediastinal lymphadenopathy in patients with or without previous malignancy: EUS-FNA-based differential cytodiagnosis in 153 patients. Am J Gastroenterol 95:2278-2284, 2000
Wiersema MJ, Vazquez-Sequeiros E, Wiersema LM: Evaluation of mediastinal lymphadenopathy with endoscopic US-guided fine-needle aspiration biopsy. Radiology 219(1):252-257, 2001
Williams DB, Sahai AV, Aabakken L, et al: Endoscopic ultrasound guided fine needle aspiration biopsy: A large single centre experience. Gut 44:720-726, 1999
Mazzone P, Jain P, Arroliga AC, et al: Bronchoscopy and needle biopsy techniques for diagnosis and staging of lung cancer. Clin Chest Med 23(1):137-158, 2002
Pieterman RM, van Putten JW, Meuzelaar JJ, et al: Preoperative staging of non-small-cell lung cancer with positron-emission tomography. N Engl J Med 343:254-261, 2000
van Tinteren H, Hoekstra OS, Smit EF, et al: Effectiveness of positron emission tomography in the preoperative assessment of patients with suspected non-small-cell lung cancer: The PLUS multicentre randomised trial. Lancet 359:1388-1393, 2002
Detterbeck FC, DeCamp MM, Jr, Kohman LJ, et al: Lung cancer. Invasive staging: The guidelines. Chest 123(1 Suppl):167S-175S, 2003
Smulders SA, Smeenk FW, Janssen-Heijnen ML, et al: Surgical mediastinal staging in daily practice. Lung Cancer 47:243-251, 2005
Kramer H, van Putten JW, Post WJ, et al: Oesophageal endoscopic ultrasound with fine needle aspiration improves and simplifies the staging of lung cancer. Thorax 59(7):596-601, 2004
Gould MK, Kuschner WG, Rydzak CE, et al: Test performance of positron emission tomography and computed tomography for mediastinal staging in patients with non-small-cell lung cancer: A meta-analysis. Ann Intern Med 139:879-892, 2003
Yasufuku K, Chiyo M, Sekine Y, et al: Real-time endobronchial ultrasound-guided transbronchial needle aspiration of mediastinal and hilar lymph nodes. Chest 126(1):122-128, 2004
Wallace MB, Ravenel J, Block MI, et al: Endoscopic ultrasound in lung cancer patients with a normal mediastinum on computed tomography. Ann Thorac Surg 77(5):1763-1768, 2004(Jouke T. Annema, Michel I)