Malignant lymphoma consists of a diverse group of malignant solid tumors of the lymphoid tissues. They can be variously derived from B-cell progenitors, T-cell progenitors, mature B cells, or mature T cells. Lymphoma is usually categorized as Hodgkin lymphoma (HL) and non-Hodgkin lymphoma (NHL), with the latter accounting for approximately 90% of cases.1 An estimated 355,900 new cases and 191,400 deaths from NHL occurred in 2008 worldwide.2
The initial evaluation of the patient with suspected malignant lymphoma is comprised by clinical presentation, including the presence of B symptoms1 and radiographic imaging.3,4 Furthermore, establishing the histologic subtype of the lymphoma and demonstrating extent and sites of disease (localized or advanced; nodal or extranodal) is essential to determine stage of disease, treatment strategies, and prognosis. Therefore, obtaining adequate pathologic specimens is essential.1,5 Both HL and NHL can present with solitary mediastinal lymphadenopathy, sometimes leading to compression of mediastinal structures and associated symptoms such as cough, chest discomfort, or even vena cava superior syndrome.1 For the diagnosis of lymphoma with mediastinal lymphadenopathy, traditionally mediastinoscopy is used by which large histologic tissue samples can be obtained. Drawbacks are the fact that it is an invasive procedure that is usually performed only once6 and has limitations when used for a mediastinal reevaluation.
Over the last years, endosonography (both esophageal endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) and endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) have been proven to be a minimally invasive and safe alternative for surgery to sample intrathoracic lymph nodes.7 Endosonography is recommended as the initial method for mediastinal staging of lung cancer as it improves regional nodal staging and reduces futile thoracotomies in comparison with mediastinoscopy.8,9 The role of endosonography in diagnosing lymphoma with mediastinal lymphadenopathy is under debate as a wide range of sensitivities (between 57% and 91%) have been reported.10–12
Traditionally, large histologic samples were needed for a classifying diagnosis of lymphoma. It is currently under discussion to what extent and under which premises endosonography is useful for diagnosing patients with suspected malignant lymphoma presenting with mediastinal lymphadenopathy, in particular in the subgroups of patients with suspected primary lymphoma versus those with recurrent lymphoma where tissue is available for comparison.
Therefore, the aim of this study was to compare the sensitivity and negative predictive value (NPV) of endosonography with FNAs of mediastinal nodes in patients with a clinical suspicion of primary versus those with recurrent lymphoma.
Patients with suspected primary and recurrent malignant lymphoma presenting with mediastinal lymphadenopathy that underwent endosonography for diagnostic purposes between 2001 and 2011 were retrospectively identified and analyzed. Patients were identified by searching the endosonography database of the Department of Pulmonology of the Leiden University Medical Center (LUMC) and the Leeuwarden Medical Center (MCL). Lymphoma was the most likely diagnosis and the suspicion was based on the clinical presentation, radiographic findings such as mediastinal lymphadenopathy on contrast enhanced computed tomography (CT) (short axis >10 mm), the absence of intrapulmonary lesions, and a history of prior lymphoma in case of suspected recurrence. Data were collected on age, sex, mediastinal lymph node enlargement on CT, number of needle passes during endosonography, outcomes of subsequent diagnostic tests and final diagnosis. Endosonography is incorporated in the routine care for patients with suspected mediastinal malignancy. For retrospective analyses of data obtained for routine procedures, no additional ethical approval was indicated.
Endosonography Procedures and Tissue Handling
The EUS-FNA and EBUS-TBNA examinations were performed in an ambulatory setting at the Department of Pulmonology at the Leiden University Medical Center and at the Medical Center Leeuwarden. For EUS, Pentax FG 34UX or 38UX echo endoscopes (Pentax GmbH, Hamburg, Germany) and for EBUS, Olympus BF-UC160F-OL or Pentax 1970 UK scopes were used. Patients were investigated under conscious sedation using midazolam intravenously. Mediastinal lymph nodes within reach of EUS or EBUS were systematically investigated and sampled under ultrasound guidance from the esophagus or trachea using 22 G needles (for EUS: Hancke/Vilmann, GIP/MEDI- Globe, or Cook Medical, for EBUS COOK Medical or Olympus). Patients were observed for 2 hours after the procedure and were instructed to contact the hospital if chest or other discomfort occurred. Aspirates were both smeared on glass slides and processed for cell block evaluation (Shandon Cell Block Preparation System; Thermo Scientific, Pittsburgh, PA) and embedded in paraffin blocks. All slides were analyzed by an experienced cytopathologist using the May-Grünwald-Giemsa. Main motive for making cell blocks is to obtain tissue for immunohistochemical analysis and categorization of tumors that otherwise may not be possible from smears only. Cytokeratins (Clone AE1/AE3 and Clone OV-TL 12/30, Dako) and CD45 (Clones 2B11 and PD7/26, Dako) were used for differential diagnosis of metastatic undifferentiated carcinomas and lymphomas. CD45 (also called leukocyte common antigen) is a marker of hematopoietic cells and a reliable indicator of the hematolymphoid nature of a tumor, whereas keratin reactivity is typical of epithelial differentiation (carcinomas and some sarcomas). Examples of these stainings on lymph node aspirates are shown in Figure 1. In this study, flow cytometry (FC) was not routinely used for immunophenotyping of the nodal aspirate.
Reference Standard and Analysis
Final diagnosis was either based on surgical-pathologic staging, conclusive cell block analysis, or by clinical and radiologic follow-up for at least 12 months. Endosonography results were considered true positive (TP) when a diagnosis of lymphoma based on the lymph node aspirates (smears and/or cell block) was confirmed by surgical procedures or by longstanding clinical and radiologic follow-up. In case an alternative diagnosis was found, which was confirmed by surgery or follow-up, this was considered a true negative (TN) test result. In case mediastinal lymph nodes turned out to be reactive on cytopathologic evaluation but a lymphoma was found at a site outside the mediastinum, this was also considered as a TN result. A false negative (FN) result occurred when a surgical procedure confirmed a malignant mediastinal lymph node which endosonography failed to detect. FN findings were categorized in detection errors (no nodal tissue obtained) and sampling errors (representative nodal tissue obtained but no malignant cells detected). Detection of atypical cells—suspicious but not conclusive for a classifying diagnosis—was also considered as an FN finding. Sensitivity, specificity, diagnostic accuracy, and predictive values of E(B)US procedures were calculated using the standard definitions. All analyses were performed using the SPSS version 16.0 (SPSS, Chicago, IL).
In this retrospective study, 49 patients were identified in whom a total of 51 procedures were performed: 43 EUS, 7 EBUS, and 1 combined EUS/EBUS. Indications for performing EUS or EBUS were suspected primary lymphoma with mediastinal lymphadenopathy (32 patients) or suspected mediastinal recurrence of previously known lymphoma (17 patients). Baseline characteristics and procedural information are found in Table 1.
An overview of final diagnosis and subclassification of lymphoma in all 49 patients is shown in Table 2. In 33 patients (67%) the final diagnosis was malignant lymphoma. Of these, 9 patients had HL and 27 had NHL. In 3 patients, a malignant lymphoma was found but because of lack of histologic tissue no further subclassification could be made. In 28 of 33 patients, the lymphoma was located in mediastinal lymph nodes within reach of EUS and/or EBUS. In 5 patients, the mediastinal lymph nodes turned out benign but a lymphoma was found on a site outside the mediastinum: in the bone marrow (n=2), an abdominal lymph node (n=1), or an inguinal lymph node (n=2). In 16 patients (33%), a final diagnosis other than lymphoma was made: small cell lung cancer (n=3), sarcoidosis (n=3), nodal metastases of non–small cell lung cancer (n=2), thymoma (n=1), mesothelioma (n=1), and reactive benign mediastinal lymphadenopathy (n=6).
A flowchart of endosonography outcomes in 32 patients with suspected primary mediastinal lymphoma is presented in Figure 2. In 11 patients, E(B)US findings positive for malignant mediastinal lymphoma were confirmed by surgical procedures (n=7) or radiologic and clinical follow-up (n= 4). In 21 patients, E(B)US outcomes turned out negative for mediastinal lymphoma. In 9 cases these outcomes turned out to be FN after surgical procedures were performed. In 4 patients, a HL was found and in 5 patients an NHL. In 12 cases, E(B)US findings negative for mediastinal lymphoma were confirmed as TN after performing surgical procedures (n=7) or clinical follow-up (n=5).
A flowchart of E(B)US procedures in 17 patients with suspected recurrence of prior lymphoma is presented in Figure 3. In 7 patients, E(B)US findings positive for malignant mediastinal lymphoma were confirmed by surgical procedures (n=2) or clinical follow-up (n=5). In 9 patients, E(B)US findings negative for mediastinal lymphoma were confirmed as TN after performing surgical procedures (n=4) or clinical follow-up (n=5). In one patient, a E(B)US outcome negative for mediastinal lymphoma was found to be an FN after surgical resection of the malignant lymph nodes proved the presence of a NHL.
An overview of all FN E(B)US results is shown in Table 3. In 3 cases, E(B)US procedures failed to obtain representative tissue from the malignant lymph node resulting in a detection error. In 5 cases, a sampling error occurred when representative lymph node tissue was acquired from the E(B)US aspirates, but no malignant lymphoid cells were found. In 2 cases, a suspicion of malignant lymphoma was raised based on E(B)US cytology, but histologic tissue was required to confirm the diagnosis of malignant lymphoma.
In 42 of 49 patients (86%), cell block material was obtained in addition to cytologic smears. In 5 of 42 patients (12%), cell block material provided a diagnosis of lymphoma, whereas cytologic smears failed to show representative lymph node tissue. Furthermore, in 9 patients immunohistochemical staining on cell block material allowed for immunophenotyping of the lymphoma.
Overall, in 39 of 49 patients (80%), a correct diagnosis was made, as 21 patients turned out to be TN and 18 patients were TP. On the basis of these results, we calculated an overall sensitivity, NPV, and diagnostic accuracy of endosonography (EUS/EBUS) for diagnosing patients with clinical suspicion of malignant lymphoma with mediastinal lymphadenopathy of 64%, 68%, and 80% respectively. In the subset of patients with suspected primary lymphoma (n=32), sensitivity, NPV, and diagnostic accuracy were 55%, 57%, and 72%, respectively. In the subset of patients (n=17) with suspected mediastinal recurrence of prior lymphoma, sensitivity, NPV, and diagnostic accuracy were 88%, 90%, and 94%, respectively.
Endosonography has a high sensitivity (88%) and NPV (90%) for diagnosing recurrent lymphoma but a very moderate value in diagnosing primary lymphoma with mediastinal lymphadenopathy (sensitivity 55% and NPV 57%). In addition, endosonography established an alternative diagnosis in 16 of 49 patients (33%).
Our findings are in concordance with the previous literature on the value of endosonography in diagnosing lymphoma where sensitivities were found ranging from 38% to 79%.11–13 In one recent study, sensitivity and NPV of EBUS-TBNA in diagnosing patients with suspected lymphoma were 90.9% and 92.9%, respectively.10 Only 1 FN occurred in this study, which is unusually low compared with other comparable studies.11,12,14–17 One possible explanation could be the relative small sample size (n=25) of this study. In our cohort, a total of 10 FNs occurred.
Importantly, in the group of patients with suspected mediastinal recurrence of a previously known lymphoma (n=17), only 1 FN finding occurred. In this specific subset of patients we found a sensitivity, NPV and diagnostic accuracy of 88%, 90%, and 94%, respectively. As the architecture of the lymphoma in these patients was previously known, E(B)US obtained lymph node aspirates providing enough tissue in most cases to determine recurrence of lymphoma. These results are in accordance with a recently published study in which EBUS accurately diagnosed 15 of 15 (100%) cases of suspected relapsed lymphoma with mediastinal lymphadenopathy.18 The results of this study and our results support the assertion that minimally invasive and safe endosonography techniques are indicated to confirm suspected relapse of lymphoma in mediastinal lymph nodes.
Furthermore, in 42 patients cell block material was obtained next to cytologic smears. In 9 patients, immunohistochemical staining of the cell block material allowed for immunophenotyping of the lymphoma. In 5 of 42 patients (12%), cell block material showed added value by providing a diagnosis of lymphoma when cytologic smears failed to show lymphoma.
In our retrospective study, FC was not used for immunophenotyping of the lymphoma in addition to routine smears and cell block analysis of the lymph node aspirates. FC is useful in establishing B-cell monoclonality in B-cell NHL. Furthermore, it allows for immunophenotyping with lymphoid markers such as CD3, CD4, CD8, CD10, CD19, CD20, and CD45.19 Ribeiro et al20 previously showed a significant increase (P=0.04) in sensitivity when FC is combined with EUS-FNA (88%) compared with EUS-FNA alone (44%). Most studies on the role of endoscopic ultrasound in diagnosing lymphoma have therefore routinely included flow-cytometric analysis of cytologic samples.11,14,21,22 However, FC is not routinely available in all hospitals. Our results are therefore probably more applicable to routine clinical practice in most countries.
Several limitations apply to this study. Firstly, its retrospective design might be a bias for patient selection. Secondly, in 9 patients with a final diagnosis of lymphoma no surgical reference standard was available. In these cases, E(B)US findings were evaluated by clinical and radiographic follow-up. Excision lymph node biopsies are considered the golden standard in the diagnosis of lymphoma.23 Diagnosing lymphoma based on E(B)US cytology alone is controversial, because of the lack of tissue architecture and in some cases a lack of adequate lymph node tissue.24 In addition, pathologists might be more comfortable establishing a diagnosis in cases with recurrent lymphoma. An optimal assessment of the value of endosonography for the diagnosis of lymphoma should include a routine surgical procedure such as video assisted thoracoscopy or mediastinoscopy to confirm endosonography findings. Thirdly, only 22 G needles were used.
An alternative to overcome the limitations of E(B)US cytology without the need for surgery is the use of a 19 G Trucut biopsy (TCB) needle to obtain a core histologic specimen. Storch et al16 performed a retrospective case study on the added value of the combination EUS-FNA and TCB compared with EUS-FNA alone in patients with mediastinal lesions. They found a higher diagnostic yield when EUS-FNA and TCB were combined, although a higher complication rate was reported. Another way of obtaining a histologic sample of mediastinal lesions in a noninvasive manner is by endoscopic ultrasound guided miniforceps biopsy (MFB). In a recent study by Herth et al,17 patients with subcarinal lesions exceeding 25 mm (short axis) were evaluated by EBUS-TBNA with a 22 G needle, a 19 G needle, and an MFB. They reported an increased diagnostic yield of MFB compared with TBNA with a 22 or 19 G needle. The increase in diagnostic yield with MFB was most significant in patients with lymphoma, showing an increase of sensitivity from 36% with a 22 G needle to 88% with MFB. Importantly, no major complications were reported in this study. However, this procedure requires multiple passes of different instruments and therefore appears impractical. Therefore, the investigators of the MFB have recently developed a new bronchoscopic sampling instrument called transbronchial needle forceps (TBNF) for the EBUS-TBNA scope. This new sampling instrument combines the characteristics of a needle with that of forceps, thereby obtaining diagnostic histologic specimens of mediastinal lymph nodes. TBNF was feasible in the majority of cases (96%) and a specific diagnosis was established in 86% of cases. No major complications occurred.25 Future research should therefore focus on novel needles that enable larger tissue sampling.
In conclusion, our results demonstrate that endosonography using 22 G needles and cellblock analysis is a useful and safe diagnostic technique to diagnose suspected recurrent malignant lymphoma as well as alternative diagnoses in patients presenting with mediastinal lymphadenopathy. However, it is not suitable to establish a primary lymphoma diagnosis in these patients, and therefore subsequent surgical confirmation (eg, mediastinoscopy) is recommended.
1. Matasar MJ, Zelenetz AD. Overview of lymphoma
diagnosis and management. Radiol Clin North Am. 2008;46:175–198vii.
2. Jemal A, Bray F, Center MM, et al.. Global cancer statistics. CA Cancer J Clin. 2011;61:69–90.
3. Ilica AT, Kocacelebi K, Savas R, et al.. Imaging of extranodal lymphoma
with PET/CT. Clin Nucl Med. 2011;36:e127–e138.
4. Cheson BD. Role of functional imaging in the management of lymphoma
. J Clin Oncol. 2011;29:1844–1854.
5. Mahadevan D, Fisher RI. Novel therapeutics for aggressive non-Hodgkin’s lymphoma
. J Clin Oncol. 2011;29:1876–1884.
6. Van SP, De WM, Hendriks J, et al.. Remediastinoscopy. J Thorac Oncol. 2007;2:365–366.
7. Tournoy KG, de RF, Vanwalleghem LR, et al.. Endoscopic ultrasound reduces surgical mediastinal staging in lung cancer: a randomized trial. Am J Respir Crit Care Med. 2008;177:531–535.
8. Annema JT, van Meerbeeck JP, Rintoul RC, et al.. Mediastinoscopy vs endosonography for mediastinal nodal staging of lung cancer: a randomized trial. JAMA. 2010;304:2245–2252.
9. Tournoy KG, Keller SM, Annema JT. Mediastinal staging of lung cancer: novel concepts. Lancet Oncol. 2012;13:e221–e229.
10. Kennedy MP, Jimenez CA, Bruzzi JF, et al.. Endobronchial ultrasound-guided transbronchial needle aspiration in the diagnosis of lymphoma
. Thorax. 2008;63:360–365.
11. Mehra M, Tamhane A, Eloubeidi MA. EUS-guided FNA combined with flow cytometry in the diagnoses of suspected or recurrent intrathoracic or retroperitoneal lymphoma
. Gastrointest Endosc. 2005;62:508–513.
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. Iqbal S, DePew ZS, Kurtin PJ, et al.. Endobronchial ultrasound and lymphoproliferative disorders: a retrospective study. Ann Thorac Surg. 2012;94:1830–1834.
14. Stacchini A, Carucci P, Pacchioni D, et al.. Diagnosis of deep-seated lymphomas by endoscopic ultrasound-guided fine needle aspiration combined with flow cytometry. Cytopathology. 2012;23:50–56.
15. Al-Haddad M, Savabi MS, Sherman S, et al.. Role of endoscopic ultrasound-guided fine-needle aspiration with flow cytometry to diagnose lymphoma
: a single center experience. J Gastroenterol Hepatol. 2009;24:1826–1833.
16. Storch I, Shah M, Thurer R, et al.. Endoscopic ultrasound-guided fine-needle aspiration and Trucut biopsy in thoracic lesions: when tissue is the issue. Surg Endosc. 2008;22:86–90.
17. 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.
18. Moonim MT, Breen R, Fields PA, et al.. Diagnosis and subtyping of de novo and relapsed mediastinal lymphomas by endobronchial ultrasound needle aspiration. Am J Respir Crit Care Med. 2013;188:1216–1223.
19. Demurtas A, Accinelli G, Pacchioni D, et al.. Utility of flow cytometry immunophenotyping in fine-needle aspirate cytologic diagnosis of non-Hodgkin lymphoma
: a series of 252 cases and review of the literature. Appl Immunohistochem Mol Morphol. 2010;18:311–322.
20. Ribeiro A, Vazquez-Sequeiros E, Wiersema LM, et al.. EUS-guided fine-needle aspiration combined with flow cytometry and immunocytochemistry in the diagnosis of lymphoma
. Gastrointest Endosc. 2001;53:485–491.
21. Ribeiro A, Pereira D, Escalon MP, et al.. EUS-guided biopsy for the diagnosis and classification of lymphoma
. Gastrointest Endosc. 2010;71:851–855.
22. Yasuda I, Goto N, Tsurumi H, et al.. Endoscopic ultrasound-guided fine needle aspiration biopsy for diagnosis of lymphoproliferative disorders: feasibility of immunohistological, flow cytometric, and cytogenetic assessments. Am J Gastroenterol. 2012;107:397–404.
23. Ravinsky E, Morales C. Diagnosis of lymphoma
by image-guided needle biopsies: fine needle aspiration biopsy, core biopsy or both? Acta Cytol. 2005;49:51–57.
24. Hehn ST, Grogan TM, Miller TP. Utility of fine-needle aspiration as a diagnostic technique in lymphoma
. J Clin Oncol. 2004;22:3046–3052.
25. Herth FJ, Schuler H, Gompelmann D, et al.. Endobronchial ultrasound-guided lymph node biopsy with transbronchial needle forceps: a pilot study. Eur Respir J. 2012;39:373–377.