Stigt, Jos A. MD*; Boers, James E. MD, PhD†; Oostdijk, Ad H. MD‡; van den Berg, Jan-Willem K. MD, PhD*; Groen, Harry J. M. MD, PhD§
Enlarged mediastinal lymph nodes may be detected incidentally on computed tomography (CT) scan of the chest and often raise the question whether further analysis is indicated. Nowadays, CT is easily accessible and made for a broader range of indications including pulmonary embolism and coronary artery imaging. Thus, incidentally enlarged mediastinal nodes will be found in incremental numbers challenging the pulmonologist to make decisions about further diagnostic approach. Incidental findings are defined as imaging abnormalities not related to the indication for which the CT is requested.
Prevalences of incidental mediastinal lymphadenopathy are reported between 0.15 and 3%, but a systematic pathologic analysis of these findings has not been described so far.1–9 Several studies describe pathologic results of mediastinal masses in patients of unknown cause.10–13 The rates of malignancy ranged in these studies from 29 to 65% with entry criteria that varied significantly. A meta-analysis showed that larger nodes have a higher chance of containing malignancy.14 The prevalence of malignancy will be lower when nodes are small as is usually the case when detected by incidence on CT. This prospective study with patients referred from eight different hospitals describes the extend, size, metabolic activity, and pathologic findings of incidentally detected mediastinal lymph nodes.
PATIENTS AND METHODS
Study Entry Criterium
Patients with at least one enlarged mediastinal lymph node (shortest diameter of ≥10 mm) were selected for the study when this incidental finding was detected on CT of the thorax made for a wide range of indications other than the analysis or staging of any neoplasm. One patient was analyzed although the shortest diameter of nodes did not reach 10 mm. Patients with a poor performance score precluding the consequences of further diagnostic analysis were excluded.
Imaging and Evaluation of Nodal Disease
Patients were referred from eight different hospitals. CT scan of the thorax was performed according to local protocols. One of the authors (J.S.) reviewed all CT scans; all visible mediastinal nodes were measured along the shortest axis; and the mediastinal position was classified according to the International Association for the Study of Lung Cancer.15 In only a limited number of patients, a positron emission tomography (PET) was performed, with fluoro-2-deoxyglucose (FDG) in the mediastinal nodes determined (GE Discovery ST PET-CT scanner; General Electric, Milwaukee, WI).
Ultrasound-Guided Biopsies and Pathologic Analysis
Mediastinal lymph nodes were approached with endoscopic ultrasound (EUS) or endobronchial ultrasound (EBUS) for a diagnostic aspiration or biopsy. EUS and EBUS were performed with Pentax ultrasound endoscopes (FG-36 UX, FG-34 UX, and EB-1970 UK; Pentax, Tokyo, Japan) with a Hitachi EUB-5500 processor (Hitachi, Tokyo, Japan). The fine needle aspiration (FNA) biopsies were performed under conscious sedation with midazolam and with local anesthesia that was sprayed in the oropharynx (lidocain 1%) and lidocain gel 20 mg/ml. Per nodal site, three to four needle passes were performed, and at least two aspirates were smeared on slides. Remaining aspirate was deposited in a fixative medium (carbowax 2% [polyethyleenglycol 20 g in ethanol 96% methylated and filled up to 1000 ml with water]) for immunohistochemical staining.
From August 2005 until December 2010, 83 of 1530 patients met the inclusion criteria of our study. They underwent EUS or EBUS for analysis of incidentally detected mediastinal lymphadenopathy and if possible FDG-PET. CT was always available.
Table 1 shows the characteristics of 83 patients (61 men and 22 women) with a median age of 59 years (range 27–87 years). All patients were referred for EUS-FNA (81) or EBUS-transbronchial needle aspiration (2) of incidentally detected lymph nodes on CT scans for a variety of indications (Table 1). Most CT scans were angio-CT scans made in the workup of pulmonary embolism (43 patients). Other indications were pulmonary infiltrative disease (5 patients), pleural disease (8 patients), parenchymatous disease (3 patients), health screening scans (2 patients), coronary angiography (2 patients), and a remaining group with other indications (20 patients).
Numbers and Sizes of Mediastinal Nodes
In our patients, nodes of all sizes were visible on CT scans in at least three levels (according to the Naruke classification), and in some patients, lymph nodes were visible on CT at all sites. The median number of sites with visible nodes was 7 (range 3–9). The median number of lymph nodes of at least 10 mm was 3 (range 0–8).
Table 1 shows the numbers of patients with different lymph node sizes at any mediastinal site. The median number of involved mediastinal sites is also described in Table 1. In 64 patients, hilar lymphadenopathy was also evident. In 19 patients, hilar node enlargement was absent. Table 2 demonstrates the nodal size, subdivided into several size categories, for specified mediastinal sites.
FDG Uptake in Mediastinal Nodes
PET scans had been performed in 29 patients. Increased FDG uptake in mediastinal nodes was demonstrated in 25 patients (inadequate aspirates, 3; reactive lymphoid change, 16; granulomatous inflammation compatible with sarcoidosis, 5; and granulomatous inflammation staining positive for mycobacteria, 1). In four patients, there was no increased FDG uptake observed (all reactive lymphoid change).
Table 2 describes the results of pathologic analysis. Cytologic aspirates were derived from 80 patients. In two patients, introduction of the endoscope failed, and in one patient, nodes were not detected with ultrasound. In four patients, the aspirates were considered inadequate for pathologic analysis.
Guided by ultrasound, subcarinal nodes were biopsied in 67 patients, the aortic window (locations 4L and 5) in 34 patients, location 8 in 2 patients, and both locations 2L and 4R in 1 patient. In 76 of 80 patients, the quality of aspirated material allowed an adequate pathologic analysis.
In 55 patients, the aspirates showed many lymphocytes and were considered reactive without a classifying diagnosis. Eighteen patients had a granulomatous inflammatory reaction compatible with sarcoidosis. One patient also had a granulomatous disorder and acid-fast mycobacteria with Ziehl-Neelson stain. Conventional culture and a polymerase chain reaction for mycobacteria were negative. One patient had a bronchogenic cyst.
In a patient with a history of breast cancer, a CT scan was performed to analyze loculated pleural fluid without pathologic evidence of a malignancy. The enlarged mediastinal nodes, not visible on normal x-ray, contained metastatic disease of her breast cancer. In a subset of 21 patients with conditions frequently associated with mediastinal lymphadenopathy (pleural disease, interstitial disease, and infiltrative disease), lymphocytes were present in 81% and a granulomatous inflammatory pattern in 14% compared with 61 and 24% for all other patients, respectively.
A classifying diagnosis was obtained with EUS in 21 of 83 patients. In the 62 patients with no classifying diagnosis, follow-up data were analyzed. This group consisted of 55 patients with adequate aspirates, 4 patients with inadequate aspirates, 2 patients in whom introduction of the echoscope failed, and 1 patient in whom no enlarged lymph nodes were detected with ultrasound.
Follow-up CT scans were available for 36 of 62 patients (58%). The scans were made after a median interval of 118 days (range 10–692 days). Decisions for follow-up were made at the discretion of patients' own treating physician. Eleven patients were referred from other hospitals and were lost to follow-up, and for 16 patients in our own practice, no follow-up CT scans were requested.
In nine patients, the lymphadenopathy had reduced spontaneously. In 24 patients, the lymphadenopathy remained unchanged, and in 2 patients, the lymphadenopathy progressed both in size and number. In both patients with progressive lymphadenopathy detected on CT scans after 20 and 23 months of follow-up, the initial EUS-FNA contained adequate samples without malignant cells. In one patient, a repeat EUS-FNA after 21 months in subcarinal lymph nodes showed malignant cells, and this patient was diagnosed with stage IV non-small cell lung cancer. In the other patient, metastatic lung cancer was demonstrated by a liver biopsy 22 months after initial EUS-FNA. In this patient with obvious progression of mediastinal lymphadenopathy, metastatic disease was not verified as a repeated EUS was not performed. Revision and comparison of CT scans suggest progression of already existing lymphadenopathy, but an evident primary tumor was absent on the initial scans. When the number and the sizes of enlarged nodes for the patients with malignancy (developed during follow-up or at initial analysis) are compared with patients without malignancy, there are no differences.
In this study, a group of 83 patients referred for suspected pulmonary malignancy were described who had in common that mediastinal lymphadenopathy was detected incidentally on CT performed for a variety of indications other than the analysis of malignancy.
The most characteristic features of these enlarged mediastinal lymph nodes is their presence in multiple Naruke stations, its variable but relatively small size, and the concomitant hilar involvement in most patients. There was one patient in whom the largest nodes measured less than 10 mm. In 46% (38/83) of patients, the largest measured nodes were between 10 and 15 mm followed by 26% (22/83) of patients with maximum nodes of 16 to 20 mm, 24% (20/83) of patients with largest nodes of 21 to 30 mm, and 2% (2/83) with nodes measuring more than 30 mm.
The size of mediastinal nodes is predictive for the final pathologic diagnosis as has been demonstrated in a study where benign nodes were significantly smaller than malignant nodes.16 However, the nodes of the patient with breast cancer metastasis in this study and the two patients who developed non-small cell lung cancer during follow-up did not differ in size from other patients.
Incidental mediastinal findings are described in screening studies for coronary artery disease, but information on extend and size of these findings is scarce.1,2,4–7 Lung cancer screening studies also report prevalences of nonpulmonary incidental findings, but details of mediastinal incidental findings are limited or absent.3,8,9
If imaging was expanded with FDG-PET, the majority of patients demonstrated increased FDG uptake in their nodes. It is debatable whether PET is indicated in patients without radiologic evidence of primary lung tumors. All but 1 of 29 patients in whom PET was performed had at least 1 lymph node measuring 10 mm or more in short-axis diameter. Remarkably, the only patient with nodes smaller than 10 mm developed lung cancer with confirmed subcarinal metastatic disease. PET offered in our study added no value to CT in incidentally detected mediastinal nodes because reactive inflammatory diseases also showed metabolic activity and hence did not discriminate from malignant disease.
In our study, the majority of incidental mediastinal findings showed a reactive inflammatory pattern on pathologic analysis, sometimes compatible with sarcoidosis as might be expected from imaging results. Granulomatous inflammatory reactions could be expected relatively more often in patients with conditions with pleural, interstitial, or infiltrative disease. In a screening study in asbestos workers (1% of patients), mediastinal lymphadenopathy was reported, and all lesions were proved to be benign.9 The size of the detected mediastinal nodes was not provided, but because this was a screening study, the lesions were presumably small in size.
In a study of 1520 lung cancer screening participants, incidental mediastinal findings (2 lymphomas) were reported, but the prevalence of incidentally detected mediastinal lymphadenopathy was not described.8
Malignancy was found in only one patient in our study. This patient had a history of breast cancer, and so, this finding was not unexpected. In all other patients, malignancy was not observed. These findings conflict with the high pretest probability of malignancy in patients with mediastinal lymphadenopathy of unknown cause in former series.10–13 Critical is how patients are selected. In a study of isolated mediastinal lymphadenopathy, EBUS-transbronchial needle aspiration resulted in a diagnosis of malignancy in 60% (33/55) of patients.12 All patients in this study had suspected lymphoma, 29% (16/55) patients had a history of malignancy, and excluded were patients with typical clinicoradiologic features of sarcoidosis. This study did not specify the size and number of mediastinal nodes. Neither were these features described in two studies reporting malignancy in 29% (40/140) and 65% (22/34) of patients.10,13 Finally, in 61% (46/75) of patients, malignancy was observed in subcarinal nodes but these were at least 25 mm.11 In a subgroup of patients analyzed for isolated mediastinal lymph nodes with typical clinicoradiologic features suggestive for sarcoidosis, this diagnosis was confirmed in 93% (26/28) of patients but 1 of 28 had malignancy.17
Our study population cannot be compared with the patients in these studies because of different entry criteria and lacking information on number and size of mediastinal nodes. Regarding our own results in patients with high rates of hilar lymphadenopathy and multiple mediastinal nodes, the number of patients with a granulomatous inflammatory reaction seems small. The difference of these results with our findings can be explained by the fact that the incidental nodes in our study were relatively small in size and mostly could not be detected with normal chest radiographs.
In the literature, there are no follow-up data of incidental mediastinal findings. In our study with a deliberate follow-up, determined by the treating physicians, follow-up CT scan was available in more than half of the patients without a classifying diagnosis of incidental mediastinal findings. Two of 62 patients (3%) developed lung cancer in nodes that had been evaluated initially with EUS-FNA. The interval of detection of lung cancer was almost 2 years in both patients. Based on the follow-up findings in this study, the need for regular monitoring of incidentally detected lymphadenopathy with CT scans is debatable.
Regarding the pathologic diagnoses in this study, the probability to detect a disease that needs treatment is very low (although an exception could be made for patients with a history of malignancy). This counterweights the low threshold to apply modern endosonographic techniques because of their favorable safety profile and high diagnostic accuracy.18,19
Incidentally detected lymphadenopathy often worries pulmonologists leading to further diagnostic procedures. This explorative study demonstrates that incidentally detected mediastinal findings are mainly a manifestation of reactive inflammatory origin in mainly multiple, slightly enlarged mediastinal lymph nodes. Our study does not support a very aggressive diagnostic approach of these nodes except for patients with known malignancy or a history of malignancy.
1. Gil BN, Ran K, Tamar G, et al. Prevalence of significant noncardiac findings on coronary multidetector computed tomography angiography in asymptomatic patients. J Comput Assist Tomogr 2007;31:1–4.
2. Haller S, Kaiser C, Buser P, et al. Coronary artery imaging with contrast-enhanced MDCT: extracardiac findings. Am J Roentgenol 2006;187:105–110.
3. Henschke CI, Lee IJ, Wu N, et al. CT screening for lung cancer: prevalence and incidence of mediastinal masses. Radiology 2006;239:586–590.
4. Hunold P, Schmermund A, Seibel RM, et al. Prevalence and clinical significance of accidental findings in electron-beam tomographic scans for coronary artery calcification. Eur Heart J 2001;22:1748–1758.
5. Jacobs PC, Mali WP, Grobbee DE, et al. Prevalence of incidental findings in computed tomographic screening of the chest: a systematic review. J Comput Assist Tomogr 2008;32:214–221.
6. Onuma Y, Tanabe K, Nakazawa G, et al. Noncardiac findings in cardiac imaging with multidetector computed tomography. J Am Coll Cardiol 2006;48:402–406.
7. Schragin JG, Weissfeld JL, Edmundowicz D, et al. Non-cardiac findings on coronary electron beam computed tomography scanning. J Thorac Imaging 2004;19:82–86.
8. Swensen SJ, Jett JR, Sloan JA, et al. Screening for lung cancer with low-dose spiral computed tomography. Am J Respir Crit Care Med 2002;165:508–513.
9. Vierikko T, Jarvenpaa R, Autti T, et al. Chest CT screening of asbestos-exposed workers: lung lesions and incidental findings. Eur Respir J 2007;29:78–84.
10. Caddy G, Conron M, Wright G, et al. The accuracy of EUS-FNA in assessing mediastinal lymphadenopathy and staging patients with NSCLC. Eur Respir J 2005;25:410–415.
11. Herth FJ, Morgan RK, Eberhardt R, et al. Endobronchial ultrasound-guided miniforceps biopsy in the biopsy of subcarinal masses in patients with low likelihood of non-small cell lung cancer. Ann Thorac Surg 2008;85:1874–1878.
12. Steinfort DP, Conron M, Tsui A, et al. Endobronchial ultrasound-guided transbronchial needle aspiration for the evaluation of suspected lymphoma. J Thorac Oncol 2010;5:804–809.
13. Yasufuku K, Nakajima T, Fujiwara T, et al. Utility of endobronchial ultrasound-guided transbronchial needle aspiration in the diagnosis of mediastinal masses of unknown etiology. Ann Thorac Surg 2011;91:831–836.
14. de Langen AJ, Raijmakers P, Riphagen I, et al. The size of mediastinal lymph nodes and its relation with metastatic involvement: a meta-analysis. Eur J Cardiothorac Surg 2006;29:26–29.
15. Rusch VW, Asamura H, Watanabe H, et al. The IASLC lung cancer staging project: a proposal for a new international lymph node map in the forthcoming seventh edition of the TNM classification for lung cancer. J Thorac Oncol 2009;4:568–577.
16. Wiersema MJ, Vazquez-Sequeiros E, Wiersema LM. Evaluation of mediastinal lymphadenopathy with endoscopic US-guided fine-needle aspiration biopsy. Radiology 2001;219:252–257.
17. Steinfort DP, Hew MJ, Irving LB. Bronchoscopic evaluation of the mediastinum using endobronchial ultrasound—a description of the first 216 cases performed at an Australian tertiary hospital. Intern Med J. In press.
18. Rintoul RC, Tournoy KG, El DH, et al. EBUS-TBNA for the clarification of PET positive intra-thoracic lymph nodes-an international multi-centre experience. J Thorac Oncol 2009;4:44–48.
19. Micames CG, McCrory DC, Pavey DA, et al. Endoscopic ultrasound-guided fine-needle aspiration for non-small cell lung cancer staging: a systematic review and metaanalysis. Chest 2007;131:539–548.