Advertisement

MAGE qPCR Improves the Sensitivity and Accuracy of EBUS-TBNA for the Detection of Lymphatic Cancer Spread

      Introduction

      Microscopic examination of histologic slides or cytologic specimens of mediastinal lymph node samples obtained by diagnostic mediastinoscopy or endobronchial ultrasound-guided fine-needle aspiration (EBUS-TBNA) is routinely used for the staging of lung cancer patients. Therefore, we explored whether the detection of tumor-associated mRNA in lymph node samples from patients with suspected lung cancer adds diagnostic accuracy to conventional histopathological staging.

      Methods

      We examined 202 lymph nodes obtained by EBUS-TBNA or mediastinoscopy from 89 patients with lung cancer. Lymph node samples from patients with nonmalignant disease were available as controls (60 samples from 31 patients). Real-time quantitative mRNA analysis was performed for melanoma antigen-A genes (MAGE-A 1–6, MAGE-A 12) using a LightCycler 480 instrument.

      Results

      MAGE transcript levels in control and cancer patients differed widely, and the 95% confidence interval served to define the threshold between negative and positive samples. MAGE 1 to 6 transcripts were detected in 35 of 122 (28.7%) lymph nodes obtained by EBUS-TBNA and 16 of 80 (20.0%) lymph nodes obtained by mediastinoscopy. MAGE 12 transcripts were detected in 10 of 122 (8.2%) lymph nodes obtained by EBUS-TBNA and 9 of 80 (11.3%) lymph nodes obtained by mediastinoscopy. Although the accuracy of histopathological diagnosis after EBUS-TBNA and mediastinoscopy was 69.6% and 84.1%, respectively, it increased to 81.2% and 86.4%, respectively, when combined with MAGE-quantitative polymerase chain reaction.

      Conclusions

      The combination of EBUS-TBNA and MAGE-quantitative polymerase chain reaction increases the accuracy of tumor cell detection to the level seen with mediastinoscopy.

      Key Words

      Lung cancer is the leading cause of tumor-related deaths worldwide. The 5-year survival rate ranges from 3% to 89% depending on the extent of tumor spread.
      • Crino L
      • Weder W
      • van Meerbeeck J
      • et al.
      Early stage and locally advanced (non-metastatic) non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis treatment and follow-up.
      Surgery remains the cornerstone of early stage non-small cell lung cancer treatment, and its indication highly depends on the accuracy of mediastinal staging. Patients with ipsilateral lymph node metastasis are treated with an multimodal approach including surgery, whereas patients with contralateral mediastinal lymph node involvement are addressed by primary radiochemotherapy or chemotherapy alone as a first-line treatment.
      • Crino L
      • Weder W
      • van Meerbeeck J
      • et al.
      Early stage and locally advanced (non-metastatic) non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis treatment and follow-up.
      ,
      • D'Addario G
      • Fruh M
      • Reck M
      • et al.
      Metastatic non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up.
      Therefore, adequate staging is highly important to define accurate treatment strategies for patients with non-small cell lung cancer.
      At present, mediastinoscopy is most often used to rule out or confirm mediastinal lymph node involvement. Overall mortality ranges from 0% to 0.08%
      • Lemaire A
      • Nikolic I
      • Petersen T
      • et al.
      Nine-year single center experience with cervical mediastinoscopy: complications and false negative rate.
      ,
      • Detterbeck FC
      • Rivera MP
      • Socinski M
      • et al.
      in large series, whereas complications occur in up to 3% of cases, primarily described as massive hemorrhage after injury to the great vessels (0.4%)
      • Park BJ
      • Flores R
      • Downey RJ
      • et al.
      Management of major hemorrhage during mediastinoscopy.
      or palsy of the left recurrent laryngeal nerve (1–2%).
      • Martin LW
      Invasive mediastinal staging for non-small-cell lung cancer.
      Hence, it is an invasive diagnostic technique and less invasive staging methods, such as the endobronchial ultrasound-guided fine-needle aspiration (EBUS-TBNA), may be indicated. EBUS-TBNA is a minimally invasive method to examine mediastinal and hilar lymph nodes with a reported sensitivity of 79 to 99%.
      • Detterbeck FC
      • Jantz MA
      • Wallace M
      • et al.
      Invasive mediastinal staging of lung cancer: ACCP evidence-based clinical practice guidelines (2nd edition).
      • Andrade RS
      • Groth SS
      • Rueth NM
      • et al.
      Evaluation of mediastinal lymph nodes with endobronchial ultrasound: the thoracic surgeon's perspective.
      • Herth FJ
      • Eberhardt R
      • Vilmann P
      • et al.
      Real-time endobronchial ultrasound guided transbronchial needle aspiration for sampling mediastinal lymph nodes.
      • Vincent BD
      • El-Bayoumi E
      • Hoffman B
      • et al.
      Real-time endobronchial ultrasound-guided transbronchial lymph node aspiration.
      • Yasufuku K
      • Chiyo M
      • Koh E
      • et al.
      Endobronchial ultrasound guided transbronchial needle aspiration for staging of lung cancer.
      However, the negative predictive value (60–99%) seems to be lower than for mediastinoscopy (80–99%), and therefore, a confirmation of negative EBUS-TBNA findings by mediastinoscopy has been recommended.
      • Groth SS
      • Andrade RS
      Endobronchial and endoscopic ultrasound-guided fine-needle aspiration: a must for thoracic surgeons.
      To improve the sensitivity of the preoperative diagnostic procedures, especially EBUS-TBNA, we looked for molecular markers to detect disseminated tumor cells in lymph node samples. Recently, we established a quantitative polymerase chain reaction (qPCR) assay to detect transcripts of cytokeratin 19, a well-known marker for epithelial cancer cells. We showed that cytokeratin 19 is detected in 100% of EBUS-TBNA lymph node samples, excluding its diagnostic use. This lack of specificity of cytokeratin 19 assays may be explained by contamination of the samples with epithelial cells from the bronchial tubes.
      • Dango S
      • Cucuruz B
      • Mayer O
      • et al.
      Detection of disseminated tumour cells in mediastinoscopic lymph node biopsies and endobronchial ultrasonography-guided transbronchial needle aspiration in patients with suspected lung cancer.
      Other highly sensitive markers such as ks1/4 (EpCam, CD326) or lunx (also known as palate, lung, and nasal epithelium carcinoma-associated gene, PLUNC) are also expressed in normal bronchial epithelium and therefore might be unsuitable for EBUS-TBNA.
      • Wallace MB
      • Block MI
      • Gillanders W
      • et al.
      Accurate molecular detection of non-small cell lung cancer metastases in mediastinal lymph nodes sampled by endoscopic ultrasound-guided needle aspiration.
      However, a highly specific class of markers to detect disseminated tumor cells are the melanoma antigens (MAGE). The MAGE proteins belong to the large family of human tumor-associated antigens recognized by T cells and are known to be expressed in a large variety of neoplasms but not in normal tissues with the notable exception of testis.
      • Kirkin AF
      • Dzhandzhugazyan KN
      • Zeuthen J
      Cancer/testis antigens: structural and immunobiological properties.
      ,
      • Van den Eynde BJ
      • van der Bruggen P
      T cell defined tumor antigens.
      Because of this they are also known as cancer-testis antigens.
      • Scanlan MJ
      • Gure AO
      • Jungbluth AA
      • et al.
      Cancer/testis antigens: an expanding family of targets for cancer immunotherapy.
      Recently, we could show that MAGE transcripts detected in bone marrow, blood, and lymph nodes are a suitable marker for the detection of disseminated tumor cells.
      • Mecklenburg I
      • Weckermann D
      • Zippelius A
      • et al.
      A multimarker real-time RT-PCR for MAGE-A gene expression allows sensitive detection and quantification of the minimal systemic tumor load in patients with localized cancer.
      • Sienel W
      • Mecklenburg I
      • Dango S
      • et al.
      Detection of MAGE-A transcripts in bone marrow is an independent prognostic factor in operable non-small-cell lung cancer.
      • Xi L
      • Coello MC
      • Litle VR
      • et al.
      A combination of molecular markers accurately detects lymph node metastasis in non-small cell lung cancer patients.
      In primary lung cancers, MAGE expression has been reported to range between 30% and 85%.
      • Sienel W
      • Mecklenburg I
      • Dango S
      • et al.
      Detection of MAGE-A transcripts in bone marrow is an independent prognostic factor in operable non-small-cell lung cancer.
      ,
      • Gotoh K
      • Yatabe Y
      • Sugiura T
      • et al.
      Frequency of MAGE-3 gene expression in HLA-A2 positive patients with non-small cell lung cancer.
      • Jungbluth AA
      • Busam KJ
      • Kolb D
      • et al.
      Expression of MAGE-antigens in normal tissues and cancer.
      • Lucas S
      • De Smet C
      • Arden KC
      • et al.
      Identification of a new MAGE gene with tumor-specific expression by representational difference analysis.
      • Weynants P
      • Lethe B
      • Brasseur F
      • et al.
      Expression of mage genes by non-small-cell lung carcinomas.
      • Yoshimatsu T
      • Yoshino I
      • Ohgami A
      • et al.
      Expression of the melanoma antigen-encoding gene in human lung cancer.
      To capture the various MAGE transcripts with high probability, we decided to use primers that amplify MAGE 1 through MAGE 6, which are highly homologous.
      • Park JW
      • Kwon TK
      • Kim IH
      • et al.
      A new strategy for the diagnosis of MAGE-expressing cancers.
      In addition, to detect MAGE 12 mRNA, which is frequently expressed in lung cancer,
      • Mecklenburg I
      • Weckermann D
      • Zippelius A
      • et al.
      A multimarker real-time RT-PCR for MAGE-A gene expression allows sensitive detection and quantification of the minimal systemic tumor load in patients with localized cancer.
      ,
      • Sienel W
      • Mecklenburg I
      • Dango S
      • et al.
      Detection of MAGE-A transcripts in bone marrow is an independent prognostic factor in operable non-small-cell lung cancer.
      but poorly amplified by the universal primers, we added MAGE 12-specific primers.
      Using this molecular biological approach, we investigated the presence of disseminated tumor cells in EBUS-TBNA samples and mediastinoscopic biopsies for a refinement of the staging procedure.

      PATIENTS AND METHODS

      Patients and Sample Preparation

      The study population consisted of 120 patients, which were screened and prospectively included in the trial. Eighty-nine patients had a primary lung cancer (clinical stages IIA–IIIB) and 31 patients presented with nonmalignant diseases (sarcoidosis, tuberculosis, anthracosilicosis, or reactive lymphadenopathy). Study population's age ranged from 33 to 88 years. All patients were treated in the Department of Thoracic Surgery at the University Medical Center Freiburg between December 2007 and May 2009 after giving informed consent. The study was approved by the Ethical Committee of the University Freiburg (ethics vote 168/04). Consecutive patients routinely underwent staging with high-resolution spiral computed tomography-scan and bronchoscopy followed by EBUS-TBNA and/or mediastinoscopy. For EBUS-TBNA, we used an Olympus ViziShot system (Olympus Ltd, Tokyo, Japan) equipped with an ultrasonic 7.5-mHz longitudinal transducer. A 21-gauge needle was used for 3 to 5 needle passes for each lymph node, as described by Herth et al.
      • Herth FJ
      • Eberhardt R
      • Vilmann P
      • et al.
      Real-time endobronchial ultrasound guided transbronchial needle aspiration for sampling mediastinal lymph nodes.
      Needles were cleaned between different lymph nodes according to a specific cleaning procedure as given by the manufacturer and then reused. The specimen was judged visually by the surgeon and then analyzed directly using liquid cytology and microscopy. In case of negative lymph node samples, patients were primarily subjected to thoracotomy including lymph node extirpation. Pathological assessment of surgically removed lymph nodes was used as a benchmark. The samples were taken from mediastinal lymph node stations 4R, 4L, and 7 by EBUS-TBNA and 2R, 2L, 4R, 4L, and 7 by diagnostic mediastinoscopy. The detection of tumor cells in ipsilateral lymph nodes classified the patients as pN2 stage, whereas detection in contralateral lymph nodes classified the patients as pN3 stage. The sixth edition of the tumor, node, metastasis classification was used.
      • Mountain CF
      Revisions in the international system for staging lung cancer.

      RNA Preparation and cDNA Synthesis

      Before isolation, mediastinoscopy samples were frozen in RNA later (QIAGEN, Hilden, Germany) at –20°C and before use were ground with a mortar before transferring them to RLT buffer (QIAGEN). EBUS-TBNA samples were directly collected in RLT buffer with β-mercaptoethanol and stored at –80°C. Subsequently, total RNA was extracted from the homogenate according to the manufacturer's protocol (RNeasy mini kit, QIAGEN). RNA concentration was measured using a spectral photometer. cDNA was prepared in 20 µl of reactions using 500 ng of total RNA. All reagents were obtained from a commercially available cDNA synthesis kit (1st strand cDNA synthesis for RT-PCR [AMV], Roche Diagnostics, Mannheim, Germany). The outcome was tested with a control—PCR for the β-actin gene. Only positive samples were further analyzed.

      Cell Lines

      The human colon cancer cell line HT29 has abundant expression of MAGE-A transcripts and was used as a positive control for the qPCR assay. The cell line was maintained in Dulbecco's modified Eagle's medium (Sigma Aldrich, St. Louis, MO) supplemented with 10% fetal bovine serum (Sigma) and 1% penicillin/streptomycin at 37°C in a humidified atmosphere containing 5% CO2. The cells were harvested when they became subconfluent.

      Quantitative PCR

      For the first round of PCR (preamplification), we used 100 µl of reactions containing 1 µl of random-primed cDNA, 10 µl of 103 PCR buffer (200 mM Tris [pH 8.4], 500 mM KCl), 2 µl of 10 mM dNTP, 3 µl of 50 mM MgCl2, 5 µl of 5 µM each of the outer primer, and 0.6 U Taq DNA polymerase (Invitrogen, Darmstadt, Germany). The cycling parameters were as follows: initial denaturation at 95°C for 5 minutes, followed by 15 cycles (20 cycles for MAGE 12) of 95°C for 30 seconds, 60°C for 45 seconds, and 72°C for 45 seconds. The final extension incubation was performed at 72°C for 10 minutes. After the first PCR, qPCR was performed in a LightCycler 480 instrument (Roche Diagnostics). Each 20 µl of qPCR contained 2 µl of the first reaction, 2 µl of 5 µM each of the inner primer and 14 µøl of SYBR Green I master (Roche Diagnostics) containing SYBR green, dNTPs, MgCl2, and reaction buffer as described in the manufacturer's data sheet. Cycling parameters were 5 minutes at 95°C for initial activation of the enzyme, 15 seconds at 95°C for denaturation, 10 seconds at 61°C for annealing, and 20 seconds at 72°C for elongation for 40 cycles. After completion of the reaction, the PCR products were subjected to a melting curve analysis spanning the temperature range from 65 to 95°C with a ramping rate of 0.1°C/sec. The specificity of the amplification was further confirmed by electrophoresis on 2% ethidium bromide-stained agarose gels.
      The combination of primers MMRP3 and MMRP4 was used for the gene expression of all MAGE 1 to 6 genes together, as described by Park et al.
      • Park JW
      • Kwon TK
      • Kim IH
      • et al.
      A new strategy for the diagnosis of MAGE-expressing cancers.
      For MAGE 12 a specific primer was used, because of its relatively high expression in lung cancer.
      • Mecklenburg I
      • Weckermann D
      • Zippelius A
      • et al.
      A multimarker real-time RT-PCR for MAGE-A gene expression allows sensitive detection and quantification of the minimal systemic tumor load in patients with localized cancer.
      The sequences of all oligonucleotide primers are listed in Supplementary Table 1. For the quantitative analysis of the reference housekeeping gene porphobilinogen desaminase (PBGD), specific primers were not added to the MAGE preamplification but only for the qPCR, that is., PBGD was measured in an effective dilution of 1:10 of the initial cDNA. The samples were each measured in triplicate, a negative control was also included for each sample, and an internal standard curve was measured in each run.

      Calculations and Statistical Analysis

      The quantification of gene expression was based on the cycle number at which the fluorescence of a sample rises above the background fluorescence (crossing point) and was calculated by a standardized algorithm of the software. Relative quantification of MAGE expression was calculated in relation to the concentration of the reference housekeeping gene PBGD. To determine PCR efficiency at different target RNA concentrations, standard curves using serial dilutions of HT29 cDNA (1:1, 1:10, 1:100, 1:200, 1:1000) were performed for each experiment. In addition, for normalization, an internal calibrator was included in each run. Investigators performing qPCR were blinded concerning the histopathological results. All statistical analyses were performed using SPSS software version 17.0 for Windows (SPSS Inc., Chicago, IL). The values of p below 0.05 were considered statistically significant.

      RESULTS

      Figure 1 depicts the study design and how we evaluated the diagnostic procedures. We focused on the reported negative predictive value of EBUS-TBNA and therefore controlled each negative result after a diagnostic algorithm. In summary, the negative findings in EBUS-TBNA and mediastinoscopy were then assessed by open surgery, which is considered the gold standard.
      Figure thumbnail gr1
      Figure 1Study design. The flow chart describes the diagnostic procedures used for a mediastinal lymph node analysis of lung cancer patients. To prevent false-negative results, patients with negative endobronchial ultrasound-guided fine-needle aspiration were subjected to mediastinoscopy, and patients with negative mediastinoscopy were treated by surgical resection.
      We first analyzed the expression of MAGE transcripts in 60 lymph node samples from 31 patients with various nonmalignant disease conditions, as several studies had reported the expression of MAGE genes in chronic nonmalignant disease.
      • Jang SJ
      • Soria JC
      • Wang L
      • et al.
      Activation of melanoma antigen tumor antigens occurs early in lung carcinogenesis.
      • Kim H
      • Kim SJ
      • Lee SH
      • et al.
      Usefulness of melanoma antigen (MAGE) gene analysis in tissue samples from percutaneous needle aspiration biopsy of suspected lung cancer lesions.
      • Liu F
      • Killian JK
      • Yang M
      • et al.
      Epigenomic alterations and gene expression profiles in respiratory epithelia exposed to cigarette smoke condensate.
      • Mecklenburg I
      • Stratakis DF
      • Huber RM
      • et al.
      Detection of melanoma antigen-a expression in sputum and bronchial lavage fluid of patients with lung cancer.
      Forty samples obtained by EBUS-TBNA and 20 lymph node samples obtained by mediastinoscopy were analyzed by qPCR. Fifty-six samples (93.3%) showed no specific expression of mRNA transcripts for MAGE 1 to 6, and 54 samples (90.0%) were free of MAGE 12 transcripts. Four samples showed low levels of MAGE 1 to 6 transcripts (relative quantification values 0.0008, 0.011, 0.060, and 0.07; two samples were from patients with sarcoidosis, one with anthracosilicosis, and one with reactive lymphadenopathy) and six samples showed MAGE 12 expression (relative quantification values 0.055, 0.061, 0.100, 0.130, 0.211, and 0.616; four samples were from patients with sarcoidosis, one with tuberculosis, and one with reactive lymphadenopathy). There was no overlap between samples, which were positive for MAGE 1 to 6 and MAGE 12. Given this low expression in control samples, we decided to define a cutoff level for the qPCR result. For this, we used a relative quantification value that we calculated from the 95% confidence interval of positive values in the control group. We obtained a threshold value of 0.091 for the MAGE 1 to 6 primers and 0.420 for the MAGE 12 primers, above which MAGE expression was considered to be cancer specific (Table 1).
      TABLE 1Expression of MAGE 1–6 and MAGE 12 in Lymph Node Samples
      MAGE 1–6 ExpressionMAGE 12 Expression
      nBefore Correction
      All samples with detected transcripts were regarded as positive, independent of the relative quantification value.
      After Correction
      Only samples with a relative quantification value above the defined cutoff level (defined by the 95% confidence interval of positive samples in the control group) of 0.091 for MAGE 1–6 and 0.420 for MAGE 12 were regarded as positive.
      Before Correction
      All samples with detected transcripts were regarded as positive, independent of the relative quantification value.
      After Correction
      Only samples with a relative quantification value above the defined cutoff level (defined by the 95% confidence interval of positive samples in the control group) of 0.091 for MAGE 1–6 and 0.420 for MAGE 12 were regarded as positive.
      Lymph nodes of control patients
       EBUS-TBNA403 (7.5%)02 (5.0%)0
        Median
      Median and range of qPCR values of positive samples are listed; if only one positive sample was detected, the relative quantification value of this sample is provided.
      0.060.80
        Range
      Median and range of qPCR values of positive samples are listed; if only one positive sample was detected, the relative quantification value of this sample is provided.
      0.011–0.070.061–0.10
       Mediastinoscopy201 (5.0%)04 (20.0%)1 (5.0%)
        Median
      Median and range of qPCR values of positive samples are listed; if only one positive sample was detected, the relative quantification value of this sample is provided.
      0.000080.170.62
        Range
      Median and range of qPCR values of positive samples are listed; if only one positive sample was detected, the relative quantification value of this sample is provided.
      0.055–0.62
      Lymph nodes ofcancer patients
       EBUS-TBNA12250 (41.0%)35 (28.7%)29 (23.8%)9 (7.4%)
        Median
      Median and range of qPCR values of positive samples are listed; if only one positive sample was detected, the relative quantification value of this sample is provided.
      0.691.360.204.55
        Range
      Median and range of qPCR values of positive samples are listed; if only one positive sample was detected, the relative quantification value of this sample is provided.
      0.0021–30.40.10–30.410–10–24.50.60–24.5
       Mediastinoscopy8031 (38.8%)16 (20.0%)42 (52.5%)9 (11.3%)
        Median
      Median and range of qPCR values of positive samples are listed; if only one positive sample was detected, the relative quantification value of this sample is provided.
      0.103.240.0583.00
        Range
      Median and range of qPCR values of positive samples are listed; if only one positive sample was detected, the relative quantification value of this sample is provided.
      0.0043–28780.10–28780.00007–25.20.57–25.2
      a All samples with detected transcripts were regarded as positive, independent of the relative quantification value.
      b Only samples with a relative quantification value above the defined cutoff level (defined by the 95% confidence interval of positive samples in the control group) of 0.091 for MAGE 1–6 and 0.420 for MAGE 12 were regarded as positive.
      c Median and range of qPCR values of positive samples are listed; if only one positive sample was detected, the relative quantification value of this sample is provided.
      Next, we assessed the expression of MAGE transcripts in 202 lymph nodes of 89 patients with lung cancer (for clinical characteristics of the patients see Table 2). One hundred twenty-two lymph node samples from 69 patients were obtained by EBUS-TBNA and 80 lymph nodes from 44 patients were collected by mediastinoscopy before operation. We detected MAGE 1 to 6 mRNA in 35 of 122 (28.7%; median, 1.36; range, 0.100–30.4) and MAGE 12 mRNA in 10 of 122 EBUS-TBNA samples (8.2%; median, 4.55; range, 0.60–24.5). In the lymph node samples obtained by mediastinoscopy, MAGE 1 to 6 mRNA was detected in 16 of 80 (20.0%; median, 3.24; range, 0.100–2878) and MAGE 12 mRNA in 9 of 80 (11.3%; median, 3.00; range, 0.570–25.2) lymph nodes. A correlation with clinicopathological parameters revealed a significant correlation of MAGE 1 to 6 or MAGE 12 transcripts in EBUS-TBNA and mediastinoscopic samples with the lymph node status (p = 0.004 and 0.033, respectively), whereas age, histology, and tumor size did not show any significant correlation (Table 2). Furthermore, we observed a highly significant positive correlation between high transcript levels of MAGE 1 to 6 and MAGE 12 levels (Figures 2A, B; Spearman's [rho] r = 0.422, p < 0.001 for EBUS-TBNA and r = 0.257, p = 0.004 for mediastinoscopy samples).
      TABLE 2Clinical Characteristics of Lung Cancer Patients and MAGE Expression
      EBUS-TBNA Patients (n = 69)
      From 24 patients both EBUS-TBNA and mediastinoscopy samples were taken, resulting in a total number of 89 lung cancer patients.
      Mediastinoscopy Patients (n = 44)
      From 24 patients both EBUS-TBNA and mediastinoscopy samples were taken, resulting in a total number of 89 lung cancer patients.
      nMAGE Expressionp
      Two-sided p values were calculated by Pearson's χ2 test to determine a correlation between clinicopathologic parameters and MAGE expression. Significant p values are in bold.
      nMAGE Expressionp
      Two-sided p values were calculated by Pearson's χ2 test to determine a correlation between clinicopathologic parameters and MAGE expression. Significant p values are in bold.
      Age (yr)
       ≤60194 (21.1%)0.576126 (50.03%)0.279
       61–70238 (34.8%)295 (17.2%)
       >70279 (33.3%)33 (100.0%)
      Histology
       Squamous cell carcinoma247 (29.2%)0.995123 (25.0%)0.577
       Adenocarcinoma299 (31.0%)206 (30.0%)
       Large cell carcinoma72 (28.6%)32 (66.6%)
       Small cell lung cancer93 (33.3%)93 (33.3%)
      Tumour size
      Pathologic stages were classified according to the international union against cancer (UICC) TNM-classification.20
       pT181 (12.5%)0.438400.531
       pT23610 (27.8%)3010 (33.3%)
       pT393 (33.3%)52 (40.0%)
       pT4167 (43.8%)52 (40.0%)
      Lymph node status
      Pathologic stages were classified according to the international union against cancer (UICC) TNM-classification.20
       pN01500.0041100.033
       pN16051 (20.0%)
       pN22913 (44.8%)198 (42.1%)
       pN3198 (42.1%)95 (55.6%)
      a From 24 patients both EBUS-TBNA and mediastinoscopy samples were taken, resulting in a total number of 89 lung cancer patients.
      b Two-sided p values were calculated by Pearson's χ2 test to determine a correlation between clinicopathologic parameters and MAGE expression. Significant p values are in bold.
      c Pathologic stages were classified according to the international union against cancer (UICC) TNM-classification.20
      Figure thumbnail gr2
      Figure 2Comparison between quantitative melanoma antigens 1–6 and melanoma antigens 12 expression. The relative quantification values of melanoma antigens 1–6 and 12 gene expression were plotted against each other on a logarithmic scale for endobronchial ultrasound-guided fine-needle aspiration samples (A, Spearman's [rho] correlation coefficient, r = 0.422; p < 0.001) and mediastinoscopy samples (B, Spearman's [rho] correlation coefficient r = 0.257; p = 0.004).
      Detection of tumor-specific MAGE expression correlated with the finding of malignant cells in histopathology for mediastinoscopy samples (p = 0.010 for MAGE 1 to 6 expression and p = 0.044 for MAGE 12 expression, Figure 3A and Table 3) and EBUS-TBNA samples (p < 0.001 for MAGE 1 to 6 expression and p = 0.048 for MAGE 12 expression, Figure 3B and Table 3).
      Figure thumbnail gr3
      Figure 3Mediastinal lymph node metastasis of a squamous carcinoma of the lung. A, Histological specimen showing compact formations of tumor cells (⋆) within a desmoplastic stroma containing few rests of lymphatic tissue (L). Hematoxylin and eosin (HE)staining, 100-fold magnification. B, Cell cluster of an endobronchial ultrasound-guided fine-needle aspiration specimen with a group of tumor cells adjacent to residual lymphocytes of the lymph node. Note intracellular keratinization in one of the tumor cells (⇓). HE staining of the liquid-based cytology, 200-fold magnification.
      TABLE 3Correlation of MAGE 1–6 and MAGE 12 Expression with Histopathologic Tumor Diagnosis
      Histopathologic Tumor DiagnosisNo. of Lymph Node Samples (%)P
      Determined by χ2 test.
      NegativePositive
      EBUS-TBNA (n = 122)
       MAGE 1–6
        Negative553287 (71.3)<0.001
        Positive62935 (28.7)
       MAGE 12
        Negative5953112 (91.8)0.048
        Positive2810 (8.2)
      Mediastinoscopy (n = 80)
       MAGE 1–6
        Negative392564 (80.0)0.010
        Positive41216 (20.0)
       MAGE 12
        Negative413071 (88.8)0.044
        Positive279 (11.3)
      Percentages add up to more than 100.0 because of rounding for one decimal.
      a Determined by χ2 test.
      b Percentages add up to more than 100.0 because of rounding for one decimal.
      The extent of regional lymph node metastasis at diagnosis is important for therapy selection of lung cancer patients. Therefore, we tested whether the combination of MAGE qPCR and histopathology more accurately stratifies lung cancer patients than histopathology alone. In total, we included 69 patients who underwent preoperative EBUS-TBNA and 44 patients from whom lymph nodes were obtained by mediastinoscopy. MAGE 1 to 6 mRNA was detected in 21 of 69 patients (30.4%) and 11 of 44 patients (25.0%), respectively. Cancer-specific MAGE 12 transcripts were detected in 8 of 69 (11.6%) and 7 of 44 (15.9%) patients. For EBUS-TBNA samples, all eight MAGE 12-positive patients also showed MAGE 1 to 6 transcripts, whereas three of seven MAGE 12-positive patients who underwent mediastinoscopy showed no MAGE 1 to 6 transcripts. All but one of the patients, who displayed cancer-specific MAGE transcript levels before operation, were postoperatively diagnosed with extensive regional lymph node metastasis (pN2–3). Given these observations, we combined MAGE qPCR with histopathology of the EBUS-TBNA and mediastinoscopy samples. We reasoned that a more sensitive preoperative diagnostic procedure might prevent nonrecommended operation of patients with extended lymph node spread (pN2 or pN3 disease). Among our patients, cytopathological evaluation of preoperative EBUS-TBNA alone correctly identified 27 of 48 patients as pN2 or pN3 (accuracy 69.6%) and histopathological evaluation of mediastinoscopy 21 of 28 patients (accuracy 84.1%), respectively. The addition of cancer-specific MAGE expression resulted in increased sensitivity in tumor cell detection, providing a correct diagnosis of pN2 or pN3 in 35 of 48 EBUS-TBNA samples (accuracy 81.2%; Table 4) and 23 of 28 mediastinoscopy samples (accuracy 86.4%; Table 4).
      TABLE 4Statistical Measures for the Applied Preoperative Diagnostic Assays for Lymph Node Staging
      EBUS-TBNAMediastinoscopy
      Sensitivity (%)Specificity (%)Accuracy (%)Sensitivity (%)Specificity (%)Accuracy (%)
      Histopathology only56.3100.069.675.0100.084.1
      MAGE qPCR only43.8100.060.946.493.863.6
      Combined assay72.1100.081.282.193.886.4

      DISCUSSION

      In this study, we investigated the impact of qPCR on the detection of disseminated tumor cell mRNA in mediastinal lymph node samples and on its accuracy when combined with conventional histopathology. We established a highly sensitive and tumor-specific real-time PCR for the quantitative assessment of disseminated tumor cells in lymph node samples of lung cancer patients.
      Among our control patients, we detected MAGE 12 transcripts in four lymph nodes from patients with sarcoidosis. Although MAGE expression has not been linked to sarcoidosis until to date, these results are in agreement with previous studies, which detected expression of MAGE mRNA in patients with chronic lung damage.
      • Jang SJ
      • Soria JC
      • Wang L
      • et al.
      Activation of melanoma antigen tumor antigens occurs early in lung carcinogenesis.
      ,
      • Mecklenburg I
      • Stratakis DF
      • Huber RM
      • et al.
      Detection of melanoma antigen-a expression in sputum and bronchial lavage fluid of patients with lung cancer.
      In our study, the relative expression of MAGE transcripts in samples from patients with a nonmalignant bronchial disease was low, and cutoff levels for cancer-specific MAGE 1 to 6 and MAGE 12 expression could be defined. By these cutoff levels based on the 95% confidence interval of control samples, we were able to clearly separate MAGE expression values from control and cancer patients. By this, our assay was designed for high specificity to exclude false-positive samples, while on the other hand resulting in a decrease of sensitivity. Therefore, we consider our MAGE qPCR assay to be additive to cytological/histological analysis of samples investigated by EBUS-TBNA or mediastinoscopy.
      In patients with primary lung cancer, we detected MAGE 1 to 6 expression in 28.7% of EBUS-TBNA and 20.0% of mediastinoscopy samples, and MAGE 12 expression in 8.2% of EBUS-TBNA and 11.3% of mediastinoscopy samples (Table 4). We could not find any correlation between MAGE gene expression and the histological tumor type, confirming the results of several previous studies.<sup>19,23–25</sup> The rather low detection rate of MAGE 12 may be related to the intratumoral heterogeneity of MAGE expression, which has also been reported for melanoma.
      • dos Santos NR
      • Torensma R
      • de Vries TJ
      • et al.
      Heterogeneous expression of the SSX cancer/testis antigens in human melanoma lesions and cell lines.
      ,
      • Jungbluth AA
      • Chen YT
      • Stockert E
      • et al.
      Immunohistochemical analysis of NY-ESO-1 antigen expression in normal and malignant human tissues.
      Differences in promoter methylation, which regulate MAGE expression on a transcriptional level in different cellular subtypes,
      • Fratta E
      • Sigalotti L
      • Colizzi F
      • et al.
      Epigenetically regulated clonal heritability of CTA expression profiles in human melanoma.
      ,
      • Sigalotti L
      • Fratta E
      • Coral S
      • et al.
      Intratumor heterogeneity of cancer/testis antigens expression in human cutaneous melanoma is methylation-regulated and functionally reverted by 5-aza-2′-deoxycytidine.
      were suggested as an underlying mechanism. For example, a recent study described MAGE transcript up-regulation by genomic hypomethylation after incubation of normal tissue with cigarette smoke concentrate for up to 9 months.
      • Liu F
      • Killian JK
      • Yang M
      • et al.
      Epigenomic alterations and gene expression profiles in respiratory epithelia exposed to cigarette smoke condensate.
      This finding supports the view that MAGE gene expression is linked to early events in carcinogenesis, especially in its role to inhibit the pro-apoptotic functions of wild-type p53.
      • Marcar L
      • Maclaine NJ
      • Hupp TR
      • et al.
      Mage-A cancer/testis antigens inhibit p53 function by blocking its interaction with chromatin.
      • Yang B
      • O'Herrin S
      • Wu J
      • et al.
      Select cancer testes antigens of the MAGE-A, -B and -C families are expressed in mast cell lines and promote cell viability in vitro and in vivo.
      • Yang B
      • O'Herrin SM
      • Wu J
      • et al.
      MAGE-A mMage-b and MAGE-C proteins form complexes with KAP1 and suppress p53-dependent apoptosis in MAGE-positive cell lines.
      Furthermore, we tested whether MAGE qPCR could increase the diagnostic accuracy of the currently used methods EBUS-TBNA and mediastinoscopy. In our hands, routine cytology of EBUS-TBNA samples was less sensitive for the detection of lymph node involvement in lung cancer patients than histopathology of mediastinoscopy samples (accuracy 69.6% and 84.1%, respectively). The addition of the quantitative MAGE qPCR assay to cytologic diagnosis improved the accuracy to 81.2% and 86.4%, respectively. The increase was more pronounced for EBUS-TBNA, resulting in a comparable accuracy as for mediastinoscopy alone. The sensitivity of EBUS-TBNA in our study is 56.3% and therefore lower compared with published literature (79–95%).
      • Detterbeck FC
      • Jantz MA
      • Wallace M
      • et al.
      Invasive mediastinal staging of lung cancer: ACCP evidence-based clinical practice guidelines (2nd edition).
      This is probably due to the relatively low number of patients and the recent establishment of EBUS-TBNA as a diagnostic procedure in our department at the time of the study. EBUS-TBNA is a challenging procedure that requires considerable expertise and training, and the reported false-negative rate in most published studies indicates that EBUS-TBNA is largely operator dependent.
      • Detterbeck FC
      • Jantz MA
      • Wallace M
      • et al.
      Invasive mediastinal staging of lung cancer: ACCP evidence-based clinical practice guidelines (2nd edition).
      However, because we show that a quantitative MAGE qPCR assay supplements EBUS-TBNA in the diagnostic setting and ensures high diagnostic accuracy of EBUS-TBNA from the beginning, it could provide significant advantages during the necessary training periods of clinicians.
      • Kemp SV
      • El Batrawy SH
      • Harrison RN
      • et al.
      Learning curves for endobronchial ultrasound using cusum analysis.
      ,
      • Navani N
      • Nankivell M
      • Nadarajan P
      • et al.
      The learning curve for EBUS-TBNA.
      Several studies indicate that detection of few disseminated cancer cells in a pN0 staged lymph node has an impact on survival.
      • Kubuschok B
      • Passlick B
      • Izbicki JR
      • et al.
      Disseminated tumor cells in lymph nodes as a determinant for survival in surgically resected non-small-cell lung cancer.
      ,
      • Osaki T
      • Oyama T
      • Gu CD
      • et al.
      Prognostic impact of micrometastatic tumor cells in the lymph nodes and bone marrow of patients with completely resected stage I non-small-cell lung cancer.
      Whether MAGE positive lymph nodes predict poor survival has not been analyzed so far and demands larger diagnostic studies. Taken together, molecular-pathological detection of MAGE transcripts in lymph nodes may be used in mediastinal staging, especially as addition to the less invasive EBUS-TBNA. As mediastinoscopy is an invasive method, we suggest future prospective studies on larger cohorts of patients to investigate whether EBUS-TBNA in combination with MAGE qPCR could replace mediastinoscopy for routine staging of lung cancer and therapy planning.

      ACKNOWLEDGMENTS

      Supported by Deutsche Krebshilfe (GZ 108433).

      REFERENCES

        • Crino L
        • Weder W
        • van Meerbeeck J
        • et al.
        Early stage and locally advanced (non-metastatic) non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis treatment and follow-up.
        Ann Oncol. 2010; 21: v103-v115
        • D'Addario G
        • Fruh M
        • Reck M
        • et al.
        Metastatic non-small-cell lung cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up.
        Ann Oncol. 2010; 21: v116-v119
        • Lemaire A
        • Nikolic I
        • Petersen T
        • et al.
        Nine-year single center experience with cervical mediastinoscopy: complications and false negative rate.
        Ann Thorac Surg. 2006; 82 (discussion 1189-1190): 1185-1189
        • Detterbeck FC
        • Rivera MP
        • Socinski M
        • et al.
        Diagnosis and Treatment of Lung Cancer: an Evidence-Based Guide for the Practicing Clinician. Saunders, Philadelphia2001
        • Park BJ
        • Flores R
        • Downey RJ
        • et al.
        Management of major hemorrhage during mediastinoscopy.
        J Thorac Cardiovasc Surg. 2003; 126: 726-731
        • Martin LW
        Invasive mediastinal staging for non-small-cell lung cancer.
        Gastrointest Endosc. 2008; 67: 199-201
        • Detterbeck FC
        • Jantz MA
        • Wallace M
        • et al.
        Invasive mediastinal staging of lung cancer: ACCP evidence-based clinical practice guidelines (2nd edition).
        Chest. 2007; 132: 202S-220S
        • Andrade RS
        • Groth SS
        • Rueth NM
        • et al.
        Evaluation of mediastinal lymph nodes with endobronchial ultrasound: the thoracic surgeon's perspective.
        J Thorac Cardiovasc Surg. 2010; 139 (discussion 582-583): 578-582
        • Herth FJ
        • Eberhardt R
        • Vilmann P
        • et al.
        Real-time endobronchial ultrasound guided transbronchial needle aspiration for sampling mediastinal lymph nodes.
        Thorax. 2006; 61: 795-798
        • Vincent BD
        • El-Bayoumi E
        • Hoffman B
        • et al.
        Real-time endobronchial ultrasound-guided transbronchial lymph node aspiration.
        Ann Thorac Surg. 2008; 85: 224-230
        • Yasufuku K
        • Chiyo M
        • Koh E
        • et al.
        Endobronchial ultrasound guided transbronchial needle aspiration for staging of lung cancer.
        Lung Cancer. 2005; 50: 347-354
        • Groth SS
        • Andrade RS
        Endobronchial and endoscopic ultrasound-guided fine-needle aspiration: a must for thoracic surgeons.
        Ann Thorac Surg. 2010; 89: S2079-S2083
        • Dango S
        • Cucuruz B
        • Mayer O
        • et al.
        Detection of disseminated tumour cells in mediastinoscopic lymph node biopsies and endobronchial ultrasonography-guided transbronchial needle aspiration in patients with suspected lung cancer.
        Lung Cancer. 2010; 68: 383-388
        • Wallace MB
        • Block MI
        • Gillanders W
        • et al.
        Accurate molecular detection of non-small cell lung cancer metastases in mediastinal lymph nodes sampled by endoscopic ultrasound-guided needle aspiration.
        Chest. 2005; 127: 430-437
        • Kirkin AF
        • Dzhandzhugazyan KN
        • Zeuthen J
        Cancer/testis antigens: structural and immunobiological properties.
        Cancer Invest. 2002; 20: 222-236
        • Van den Eynde BJ
        • van der Bruggen P
        T cell defined tumor antigens.
        Curr Opin Immunol. 1997; 9: 684-693
        • Scanlan MJ
        • Gure AO
        • Jungbluth AA
        • et al.
        Cancer/testis antigens: an expanding family of targets for cancer immunotherapy.
        Immunol Rev. 2002; 188: 22-32
        • Mecklenburg I
        • Weckermann D
        • Zippelius A
        • et al.
        A multimarker real-time RT-PCR for MAGE-A gene expression allows sensitive detection and quantification of the minimal systemic tumor load in patients with localized cancer.
        J Immunol Methods. 2007; 323: 180-193
        • Sienel W
        • Mecklenburg I
        • Dango S
        • et al.
        Detection of MAGE-A transcripts in bone marrow is an independent prognostic factor in operable non-small-cell lung cancer.
        Clin Cancer Res. 2007; 13: 3840-3847
        • Xi L
        • Coello MC
        • Litle VR
        • et al.
        A combination of molecular markers accurately detects lymph node metastasis in non-small cell lung cancer patients.
        Clin Cancer Res. 2006; 12: 2484-2491
        • Gotoh K
        • Yatabe Y
        • Sugiura T
        • et al.
        Frequency of MAGE-3 gene expression in HLA-A2 positive patients with non-small cell lung cancer.
        Lung Cancer. 1998; 20: 117-125
        • Jungbluth AA
        • Busam KJ
        • Kolb D
        • et al.
        Expression of MAGE-antigens in normal tissues and cancer.
        Int J Cancer. 2000; 85: 460-465
        • Lucas S
        • De Smet C
        • Arden KC
        • et al.
        Identification of a new MAGE gene with tumor-specific expression by representational difference analysis.
        Cancer Res. 1998; 58: 743-752
        • Weynants P
        • Lethe B
        • Brasseur F
        • et al.
        Expression of mage genes by non-small-cell lung carcinomas.
        Int J Cancer. 1994; 56: 826-829
        • Yoshimatsu T
        • Yoshino I
        • Ohgami A
        • et al.
        Expression of the melanoma antigen-encoding gene in human lung cancer.
        J Surg Oncol. 1998; 67: 126-129
        • Park JW
        • Kwon TK
        • Kim IH
        • et al.
        A new strategy for the diagnosis of MAGE-expressing cancers.
        J Immunol Methods. 2002; 266: 79-86
        • Mountain CF
        Revisions in the international system for staging lung cancer.
        Chest. 1997; 111: 1710-1717
        • Jang SJ
        • Soria JC
        • Wang L
        • et al.
        Activation of melanoma antigen tumor antigens occurs early in lung carcinogenesis.
        Cancer Res. 2001; 61: 7959-7963
        • Kim H
        • Kim SJ
        • Lee SH
        • et al.
        Usefulness of melanoma antigen (MAGE) gene analysis in tissue samples from percutaneous needle aspiration biopsy of suspected lung cancer lesions.
        Lung Cancer. 2010; 69: 284-288
        • Liu F
        • Killian JK
        • Yang M
        • et al.
        Epigenomic alterations and gene expression profiles in respiratory epithelia exposed to cigarette smoke condensate.
        Oncogene. 2010; 29: 3650-3664
        • Mecklenburg I
        • Stratakis DF
        • Huber RM
        • et al.
        Detection of melanoma antigen-a expression in sputum and bronchial lavage fluid of patients with lung cancer.
        Chest. 2004; 125: 164S-166S
        • dos Santos NR
        • Torensma R
        • de Vries TJ
        • et al.
        Heterogeneous expression of the SSX cancer/testis antigens in human melanoma lesions and cell lines.
        Cancer Res. 2000; 60: 1654-1662
        • Jungbluth AA
        • Chen YT
        • Stockert E
        • et al.
        Immunohistochemical analysis of NY-ESO-1 antigen expression in normal and malignant human tissues.
        Int J Cancer. 2001; 92: 856-860
        • Fratta E
        • Sigalotti L
        • Colizzi F
        • et al.
        Epigenetically regulated clonal heritability of CTA expression profiles in human melanoma.
        J Cell Physiol. 2010; 223: 352-358
        • Sigalotti L
        • Fratta E
        • Coral S
        • et al.
        Intratumor heterogeneity of cancer/testis antigens expression in human cutaneous melanoma is methylation-regulated and functionally reverted by 5-aza-2′-deoxycytidine.
        Cancer Res. 2004; 64: 9167-9171
        • Marcar L
        • Maclaine NJ
        • Hupp TR
        • et al.
        Mage-A cancer/testis antigens inhibit p53 function by blocking its interaction with chromatin.
        Cancer Res. 2010; 70: 10362-10370
        • Yang B
        • O'Herrin S
        • Wu J
        • et al.
        Select cancer testes antigens of the MAGE-A, -B and -C families are expressed in mast cell lines and promote cell viability in vitro and in vivo.
        J Invest Dermatol. 2007; 127: 267-275
        • Yang B
        • O'Herrin SM
        • Wu J
        • et al.
        MAGE-A mMage-b and MAGE-C proteins form complexes with KAP1 and suppress p53-dependent apoptosis in MAGE-positive cell lines.
        Cancer Res. 2007; 67: 9954-9962
        • Kemp SV
        • El Batrawy SH
        • Harrison RN
        • et al.
        Learning curves for endobronchial ultrasound using cusum analysis.
        Thorax. 2010; 65: 534-538
        • Navani N
        • Nankivell M
        • Nadarajan P
        • et al.
        The learning curve for EBUS-TBNA.
        Thorax. 2011; 66: 352-353
        • Kubuschok B
        • Passlick B
        • Izbicki JR
        • et al.
        Disseminated tumor cells in lymph nodes as a determinant for survival in surgically resected non-small-cell lung cancer.
        J Clin Oncol. 1999; 17: 19-24
        • Osaki T
        • Oyama T
        • Gu CD
        • et al.
        Prognostic impact of micrometastatic tumor cells in the lymph nodes and bone marrow of patients with completely resected stage I non-small-cell lung cancer.
        J Clin Oncol. 2002; 20: 2930-2936