Diagnostic value of MUC1 and EpCAM mrna as tumor markers in differentiating benign from malignant pleural effusion

Transcription

1 Q J Med 2014; 107: doi: /qjmed/hcu130 Advance Access Publication 17 June 2014 Diagnostic value of MUC1 and EpCAM mrna as tumor markers in differentiating benign from malignant pleural effusion W. SUN 1,J.LI 1, H.-G. JIANG 1, L.-P. GE 1 and Y. WANG 2 From the 1 Department of Pulmonary Medicine and 2 Center of Experimental Medicine, Affiliated Hospital of Jiangsu University, 438 North Jiefang Street, Zhenjiang , China Address correspondence to Prof. J. Li, Department of Pulmonary Medicine, Affiliated Hospital of Jiangsu University, 438 North Jiefang Street, Zhenjiang , China. lijian541226@163.com Received 16 April 2014 and in revised form 21 May 2014 Summary Background: The sensitivity of conventional cytology for the detection of tumor cells in pleural effusion (PE) is inadequate. Mucine 1 (MUC1) and epithelial cell adhesion molecule (EpCAM) are two frequently and intensely expressed tumor-associated antigens in malignancies of epithelial origin. Objective: To evaluate the diagnostic value of pleural fluid MUC1 and EpCAM mrna in differentiating benign and malignant PE. Method: Fifty-eight patients with malignant PE and 40 patients with benign PE were included in this study. Pleural fluid MUC1 and EpCAM mrna levels were measured by quantitative real-time PCR. Carcinoembryonic antigen (CEA) and cytokeratin 19 fragment (Cyfra21-1) were also detected simultaneously. The receiver operating characteristic (ROC) curves were constructed to assess diagnostic performance of the four tumor markers in PE. Results: For the diagnosis of malignant PE, MUC1 and EpCAM mrna had larger area under ROC curves (0.916 and 0.922) and higher sensitivity (67.2 and 70.7%) with the same specificity, when compared with CEA and Cyfra21-1 (0.821 and 0.780; 48.3 and 44.8%, respectively). By combining cytology with MUC and EpCAM, a positive result indicating the presence of malignancy was achieved in 87.9%, with a good specificity of 95%. Conclusions: Compared with CEA and Cyfra21-1, the performance of MUC1 and EpCAM mrna in malignant PE diagnosis was better. MUC1 and EpCAM mrna in combination with cytology is a highly sensitive and specific diagnostic tool for detecting malignancy in PE. Introduction Malignancy is the most common cause of exudative pleural effusion (PE) and the majority of malignant PEs are due to metastatic disease. 1,2 Differentiation between malignant and benign PE has particular clinical relevance, 3,4 and more than half of the patients with malignant PE died within 6 months of the diagnosis of PE. 5 Early diagnosis and management of malignant PE may improve outcomes. The detection of tumor cells in PE is essential to establish a diagnosis. However, the detection of malignancy in PE by conventional cytology is hampered by the problem of differentiating malignant cells from reactive mesothelial cells, cytology examination only identify tumor cells in 40% of malignant PE. 6 Blind needle pleural biopsy confers little additive diagnostic value for false-negative cytology, 2 while ultrasound guided pleural biopsy can improve the diagnostic yield. 7 To further improve the diagnosis of malignant PE, many studies have investigated the use of different tumor markers in pleural fluid, the best known of which are carcinoembryonic antigen (CEA) and Cyfra21-1, with diagnostic accuracy! The Author Published by Oxford University Press on behalf of the Association of Physicians. All rights reserved. For Permissions, please journals.permissions@oup.com

2 1002 W. Sun et al. higher than that of other tumor markers However, false-positive results of the two tumor markers may weaken their diagnostic value 16 ; thus, there is no consensus of these results with regard to which markers are most effective at differentiating malignant from benign PE. Mucine 1 (MUC1) is a cell surface glycoprotein and expressed at a based level by normal ductal epithelial cells of secretory organs. 17 As a tumorassociated antigen, MUC1 is overexpressed on various carcinomas of epithelial origin, including lung, breast, ovary and colon cancer. 18 Studies have demonstrated that an elevated level of MUC1 protein plays a role in tumor progression, especially in the process of metastasis. 19 Epithelial cell adhesion molecule (EpCAM) is also a transmembrane glycoprotein that is expressed on the basolateral surface of most normal epithelial tissue. 20 Likewise, EpCAM can now be considered to be one of the most frequently and most intensely expressed tumorassociated antigens known. It is overexpressed to varying digress in most human carcinomas. 20,21 The diagnostic and prognostic characteristics of EpCAM have been reported by many investigators Recently, we have used quantitative real-time PCR (QRT-PCR) method to detect the expression of MUC1 and EpCAM mrna in peripheral blood of patients with non-small cell lung cancer (NSCLC) and have demonstrated that blood MUC1 and EpCAM mrna are predictive and prognostic markers in NSCLC patients who undergo curative surgery. 23 MUC1 and EpCAM are considered as two oncogenes and involved in intercellular adhesion, proliferation, survival, carcinogenesis and metastasis formation Because MUC1 and EpCAM genes are thought to be restricted to epithelial cells, they have been used for the detection of micrometastatic tumor cell in blood, bone marrow and lymph node of patients with numerous solid tumors However, relatively few studies have been undertaken to investigate the diagnostic utility of pleural fluid MUC1 and EpCAM mrna in patients with malignant PE. 28,29 In this study, we evaluated the efficiency of pleural fluid MUC1 and EpCAM mrna QRT-PCR assay in identification of the origin of PE and examined these test as potentially useful in assisting the diagnosis of malignant PE. Method Study population From July 2011 to December 2013, 98 patients with PE who admitted to our hospital were enrolled in the study. The subjects included 60 men and 38 women, and had a mean age of 61 years. Malignant PEs were presented in 58 patients (59%), while PEs were benign in 40 patients (41%). All PEs had definite etiology documented by physical examination, chest radiology, pleural fluid biochemistry, microbiology and cytology analyses, pleural biopsy, percutaneous needle biopsy or endoscopy biopsy. When initial thoracentesis and cytology were non-diagnostic for suspected malignant PE, repeated thoracenteses and cytology or pleural biopsy were pursued as clinical dictated. All patients with suspicious pleural fluid cytology were defined as cytologic negativity. Malignant PEs were diagnosed by either pleural fluid cytology positivity or malignant cells identified in a pleural biopsy specimen. The study was proved by the Ethics Committee of the Affiliated Hospital of Jiangsu University and written informed consent was obtained from all participants. Specimen processing Pleural fluids were collected by diagnostic thoracentesis before the patients received any therapy. Each fluid specimen was centrifuged in 50-ml tubes at 1500 rpm for 10 min at 48C, Then, cell pellets were pooled together, washed once and centrifuged again at 1500 rpm for 10 min at 48C. The cell pellets were stored at 808C until use. Meanwhile, PE specimens were taken to the clinical laboratory for routine biochemical tests and carcinoembrgynic antigen (CEA) and cytokeratin 19 fragment (Cyfra21-1) assay, and pathologic department for cytology examination. Quantitative real-time PCR Total RNA was extracted from the cell pellets using TRIZOL Reagent (Invitrogen, Carslbad, CA) following the manufacturer s instructions. The RNA concentration was measured by UV absorption at 260 and 28 nm. The A260/A280 ratio was calculated to assess RNA quality and purity. First-strand cdna was produced from total RNA by using an RNA PCR Kit version 3.0 (TakaRa Bio Inc., Tokyo, Japan), according to manufacturer s instructions. QRT-PCR of the target MUC1 and EpCAM mrna and b-actin as internal control was performed on an ABI 7500 thermal cycler real-time PCR system (Applied Biosystems, Foster City, CA), using the SYBR-Green I Chemistry. Amplification primers of the two genes were synthesized by Shanghai Invitrogen Biotechnology Co, Ltd, in China. The primer sequences, annealing temperatures and expected length of synthesis fragment were shown in Table 1. The cycling conditions were as follows: a denaturation step at 958C for 30 s,

3 Pleural fluid MUC1 and EpCAM mrna 1003 Table 1 Sequences and features of the primers used for QRT-PCR mrna Primer sequence ( ) Annealing temperature followed by 40 cycles of 958C for 5 s, annealing for 30 s at 578C and elongation for 30 s at 728C. Detection of PCR products was accomplished by measuring the emitting fluorescence (Rn) at the end of each reaction step (reaction cycles). Threshold cycle (Ct) correspond with the cycle number required to detect a fluorescence signal above the baseline. Calculations were made with the use of the comparative Ct (2 Ct ) method. 30 Each experiment was carried out in triplicate, the average value of the replicates was used as quantitative value for each specimen. All measurements were performed without any knowledge of the patients diagnostic information. Pleural fluid CEA and Cyfra21-1 were detected using chemiluminescent-microparticle immunoassay (Architect i2oo SR., Abbott, USA) technology with a commercial reagent Kit (Beckman Coulter, Inc., USA) following the manufacturer s instructions. Statistical analysis Length (bp) MUC1 AATGAATGGCTCAAAACTTGG 568C 203 CAGTAGGTTCTCACTCGCTCAG EpCAM AGTTGTTAGGGTTCTGGTCAT 588C 213 GCTCTTGGTAGTCGGTGCT b-actin TGACGTGGACATCCGCAAAG 608C 205 CTGGAAGGTGGACAGCGAGG Statistical analysis was performed with SPSS software 13.0 (SPSS Inc., Chicago, IL). Data were expressed as the median (min max) for each tumor marker because of its non-normally distributed variable. Comparisons between groups used the Mann- Whitney s U-test for continuous variable, and the Fisher s exact test for categorical variable. The correlation between the presence of malignancy and all variable was determined by logistic regression. The diagnostic performance of each tumor marker to distinguish malignant from benign PE was evaluated with the receiver operating characteristic (ROC) curve analysis. According to the optimum cutoff point provided by ROC curve analysis, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and accuracy were calculated. Statistical significance was defined by a P value <0.05. Results MUC1 and EpCAM mrna levels in malignant and benign PE The distributions of MUC1 and EpCAM mrna levels in malignant and benign PEs are represented in Figure 1. As shown in Figure 1A and B, the median levels of MUC1 and EpCAM mrna were significantly higher compared with the value in benign PE (both P < 0.001). Similarly, the median levels of CEA and Cyfra21-1 were significantly higher in malignant PE than in benign PE (P = and P = 0.003, respectively; Figure 1C and D). Levels of four tumor markers in the whole population of PE specimens are detailed in Table 2. No significant difference was found in the levels of pleural fluid MUC1 and EpCAM mrna between malignant PEs caused by lung cancer and by other cancers. The etiology of 58 malignant PEs included lung cancer (40), breast cancer (7), colon cancer (4), ovarian cancer (4) and gastric cancer (3). PEs from 40 patients with benign disease caused by tuberculous pleurisy (30), parapneumonic effusion (6) and cardiac failure (4) served as controls. Among the 58 patients with malignant PE, 22 (38%) had cytologically positive results at initial thoracentesis. In the remaining 36 patients (62%), the malignancy was confirmed finally by repeated cytologic examinations (n = 23) or blind needle biopsy (n = 13). In patients with initial cytology-negative malignant PE, the levels of MCU1 and EpCAM mrna were significantly higher compared with those of patients with benign PE (both P < 0.001, data not shown). Correlation between malignant PE and tumor markers Logistic regression analysis showed that all four tumor markers in pleural fluid, including MUC1, EpCAM, CEA and Cyfra21-1, were significantly correlated with the presence of malignant PE (Table 3). Diagnostic performance MUC1 and EpCAM mrna for differentiating malignant from benign PE The area under the ROC curve (AURC) was calculated to assess the performance of each tumor marker in diagnosis of malignant PE (Figure 2). The AURC of EpCAM mrna (0.922) appeared to be large than the AURC of MUC1 (0.916), which were greater than those of CEA and Cyfra21-1 (0.821 and 0.780, respectively). According to the ROC curves, the optimum cutoff points for

4 1004 W. Sun et al. Figure 1. Distribution of pleural fluid MUC1 mrna (A), EpCMA mrna (B), CEA (C) and Cyfra21-1 (D) levels in malignant and benign pleural effusions. Table 2 Levels of the four tumor markers in pleural effusions caused by different diseases Etiology No. of patients MUC1 mrna EpCAM mrna CEA (mg/l) Cyfra21-1 (mg/l)) Malignant 58 (59%) 8.11 ( ) ( ) 7.75 ( ) ( ) Lung ( ) ( ) 7.92 ( ) ( ) Breast ( ) ( ) 8.34 ( ) ( ) Colon ( ) ( ) 9.36 ( ) ( ) Ovarian ( ) ( ) 8.19 ( ) ( ) Gaster ( ) ( ) ( ) 9.06 ( ) Benign 40 (41%) 2.85 ( ) 4.26 ( ) 2.95 ( ) 7.15 ( ) Tuberculosis ( ) 4.26 ( ) 2.95 ( ) 7.15 ( ) Pneumonia ( ) 5.31 ( ) 3.47 ( ) 6.74 ( ) Cardiac failure ( ) 3.12 ( ) 2.14 ( ) 4.18 ( ) diagnosing malignant PE, corresponding to a diagnostic specificity of 95%, were 4.84 for MUC1, 8.65 for EpCAM, 7.72 for CEA and for Cyfra21-1, respectively. The sensitivity, specificity, PPV, NPV and accuracy of cytology and four tumor markers in pleural fluids, alone or combination in diagnosis of malignant PE, were represented in Table 4. The sensitivities of MUC1 and EpCAM mrna were markedly higher than those of CEA and Cyfra21-1, with the same specificity, although when MUC1

5 Pleural fluid MUC1 and EpCAM mrna 1005 Table 3 Logistic regression analysis Variable OR 95% CI Coefficient SE P value MUC1 mrna EpCAM mrna CEA Cyfra OR, odds ratio; CI, confidence interval; SE, standard error. cytology-negative malignant PE were positive with at least one tumor marker. These four tumor markers brought additional diagnostic information but MUC1 and EpCAM appeared to be more contributive than CEA and Cyfra21-1 as the results of their performance with ROC curve analysis, and more positive patients. Figure 2. Receive operating characteristic (ROC) curves of MUC1 mrna, EpCAM mrna, CEA and Cyfra21-1 for differentiating malignant from benign pleural effusion. The area under the ROC curves of pleural fluid MUC1 mrna, EpCAM mrna, CEA and Cyfra21-1 were (95% confidence interval [CI], ), (95% CI, ), (95% CI, ) and (95% CI, ), respectively. compared with CEA, difference did not reach statistical significance (P = 0.06). When combinations of tumor markers were evaluated, MUC1 plus EpCAM achieved a higher diagnostic value, with a sensitivity of 79.3%, specificity of 95% and accuracy of 85.7%. Whereas the combination of cytology with MUC1 and EpCAM mrna yielded a sensitivity of 87.9% with a specificity of 95% for diagnosing malignant PE. The sensitivity of the combined test is significantly elevated than that of cytology alone. In addition, we evaluated diagnostic value of MUC1 and EpEAM in patients with initial cytology-negative malignant PE. Table 5 shows the positive results of tumor markers, either individually or in combination, among the malignant PE patients with negative cytology findings in pleural fluid. Overall, 30 of 36 patients with initial Discussion In this study, we investigated the expression levels of MUC1 and EpCAM mrna in pleural fluid of patients with malignant and benign PE. Pleural fluid MUC1 and EpCAM mrna levels were significantly elevated in patients with malignant PE relative to benign pleural disease. Given the sharp differential expression levels of MUC1 and EpCAM mrna in malignant and benign PE, its detection appears to be a suitable tool for the diagnosis of malignant PE. Our results showed that pleural fluid MUC1 and EpCAM mrna contributed valuable additional information to pleural fluid cytology alone, especially when the latter was inconclusive or suspicious, in which case MUC1 and EpCAM mrna offer 95% specificity in establishing the diagnosis of malignant PE. We compared the usefulness of MUC1 and EpCAM QRT-PCR detection with cytologic examination. Frequently, MUC1 and EpCAM mrna were positive in the pleural fluid of patients exhibiting malignant cytology. However, the most striking fact was the demonstration of MUC1 and EpCAM mrna positivity in 16 and 18 of 36 cytologically negative malignant PE. In fact, the use of cytology or MRC1 and EpCAM mrna or both, made the diagnosis of 51 of 58 malignant PE (87.9%). Thus, QRT-PCR analysis of MUC1 and EpCAM mrna provide noninvasive assay sensitive enough to help physicians when cytology test remains negative. It also constitutes a good argument for physicians to rapidly carry out another thoracentesis or pleural biopsy. Detection of MUC1 and EpCAM mrna for the diagnosis of malignant PE are based on the fact that MUC1 and EpCAM have significantly

6 1006 W. Sun et al. Table 4 Diagnostic value of cytology and tumor markers for malignant pleural effusion Variable Sensitivity (%) Specificity (%) PPV (%) NPV (%) Accuracy (%) Cytology 22/58 (37.9) 40/40 (100) 22/22 (100) 40/76 (52.6) 62/98 (63.3) MUC1 39/58 (67.2) *,y 38/40 (95) 39/41 (95.1) 38/57 (66.7) 77/98 (78.6) EpCAM 41/58 (70.7) z; 38/40 (95) 41/43 (95.3) 38/55 (69.1) 79/98 (80.6) CEA 28/58 (48.3) 38/40 (95) 28/30 (93.3) 38/68 (55.9) 66/98 (67.3) Cyfra /58 (44.8) 38/40 (95) 26/28 (92.9) 38/70 (54.3) 64/98 (65.3) MUC1 + EpCAM 46/58 (79.3) 38/40 (95) 46/48 (95.8) 38/50 (76) 84/98 (65.3) CEA + MUC1 + EpCAM 38/58 (65.5) 37/40 (92.5) 38/41 (92.7) 37/57 (64.9) 75/98 (76.5) Cytology + MUC1 + EpCAM 51/58 (87.9) 38/40 (95) 51/53 (96.2) 38/45 (84.4) 89/98 (90.8) Cytology + CEA + Cyfra /58 (70.7) 38/40 (95) 41/43 (95.3) 38/55 (69.1) 79/98 (80.6) *P =0.06, compared with CEA; y P =0.024, compared with Cyfra21-1; z P =0.022, compared with CEA; P =0.008, compared with CYfra21-1. Table 5 Cases showing tumor marker positive among 36 patients with initial cytology-negative malignant PE Tumor type No. of patients Patient number with tumor marker positive MUC1 EpCAM CEA Cyfra21-1 Combination of tumor markers Lung Breast Colon Ovarian Gaster Total differential expression between carcinomas and normal adult cells, with very low levels in normal adult tissue but increased levels in wide variety of epithelial tumor The high MUC1 and EpCAM mrna levels in pleural fluid may be because of a direct shedding of MUC1 and EpCAM mrna from tumor cells into PE and, therefore, suggesting that MUC1 and EpCAM mrna may be valuable diagnostic markers for malignant PE. A previous study indicated that MUC1 mrna in pleural fluid can be applied as an adjuvant diagnostic tool for malignant PE, with sensitivity and specificity of 64.6 and 95.7%, respectively, 29 which is in agreement with ours. A study by Passebosc-Faure et al. 30 showed that detection of EpCAM mrna in pleural fluids can be used to make the diagnosis of malignant PE with 66.2% sensitivity and 100% specificity, which is also similar to our results. Meanwhile, Passebosc- Faure et al. 30 demonstrated that a combination of cytology and QRT-PCR analysis of EpCAM mrna can significantly improve the detection sensitivity of tumor cells in effusion specimens. Although CEA and Cyfra21-1 are two of the most reliable tumor markers in diagnosis of malignant PE, when compared with MUC1 and EpCAM mrna assessed in this study, we found that the latter had higher sensitivity to detect malignancies and had a larger area under ROC curves. In addition to their good performance in malignant PE diagnosis, MUC1 and EpCAM mrna are also indicators for aggressive tumor biology, so they have the potential to be two ideal tumor markers in PE diagnosis and good biomarkers in malignant PE-inducing cancer. One of the strategies for improving the sensitivity and specificity in the diagnosis of malignant PE is to search for best combination of genes for mrna analysis and other diagnostic tests including cytology or tumor markers. In this study, the combination of cytology and MUC1 plus EpCAM mrna achieved the highest sensitivity (87.9%) with a very good specificity (95%), indicating that the combination is superior to combination of CEA, Cyfra21-1 and cytology. Additionally, we excluded PEs from patients with cancer who lack confirmation of malignant PE by either pleural fluid cytologic findings or tumor cells in a pleural biopsy specimen because patients with advanced stage cancer may develop nonmalignant PE from cardiac failure, pneumonia,

7 Pleural fluid MUC1 and EpCAM mrna 1007 hypoalbuminemia, or adverse effects of anticancer drugs. Thus, unnecessary false negativity can be avoided. In conclusion, QRT-PCR assay of MUC1 and EpCAM mrna in pleural fluid may be a complementary tool for diagnosing malignant PE, especially when cytologic examination is not definitive, and can improve clinical practice. The combined use of pleural fluid cytology and MUC1 and EpCAM mrna can give a higher diagnostic performance and warrants further validation. Conflict of interest: None declared. References 1. Heffner JE. Diagnosis and management of malignant pleural effusions. Respirology 2008; 13: Specter M, Polak JS. Management of malignant pleural effusions. Semin Respir Crit Care Med 2008; 29: Light RW. Pleural effusion. N Engl J Med 2002; 346: Ferrer J, Roldan J, Teixidor J, Pallisa E, Gich I, Morell F, et al. Predictors of pleural malignancy in patients with pleural effusion undergoing thoracoscopy. Chest 2005; 127: Alemán C, Sanchez T, Alegre J, Alegre J, Ruiz E, Vázquez A, et al. Differentiating between malignant and idiopathic pleural effusion: the value of diagnostic procedure. Q J Med 2007; 100: Motherby H, Nadjari B, Friegel P, Kohaus J, Ramp U, Böcking A. Diagnostic accuracy of effusion cytology. Diagn Cytopathol 1999; 20: Koegelenberg CFN, Diacon AH. Pleural controversy: closed needle pleural biopsy or thoracoscopy which first? Respirology 2011; 16: Porcel JM, Vives M, Esquerda A, Salud A, Pérez B, Rodriguez-Panadero F. Use of a panel of tumor markers (carcinoembryonic antigen, cancer antigen 125, carbohydrate antigen 15-3, and cytokeratin 19 fragment) in pleural fluid for the differential diagnosis of benign and malignant effusions. Chest 2004; 126: Shitrit D, Zingerman B, Shitrit AB-G, Shlomi D, Kramer MR. Diagnostic value of Cyfra21-1, CEA, CA19-9, CA15-3, and CA125 assays in pleural effusions: analysis of 116 cases and review of the literature. Oncologist 2005; 10: Liang Q-L, Shi H-Z, Qin X-J, Liang X-D, Liang J, Yang H-B. Diagnostic accuracy of tumor markers for malignant pleural effusion: a meta-analysis. Thorax 2008; 63: Huang W-W, Tsao S-M, Lai C-L, Su C-C, Tseng C-E. Diagnostic value of Her-2/neu, Cyfra21-1, and carcinoembryonic antigen levels in malignant pleural effusions of lung adenocarcinoma. Pathology 2010; 42: Zhou W-B, Bai M, Jin Y. Diagnostic value of vascular endothelial growth factor and endostatin in malignant pleural effusions. Int J Tuberc Lung Dis 2009; 13: Sriram KB, Relan V, Clarke BE, Duhig EE, Yang IA, Bowman RV, et al. Diagnostic molecular biomarkers for malignant pleural effusions. Future Oncol 2011; 6: Li J, Li Z-N, Bao Q-L, Ge L-P, Li X-Q, Chen P. Evaluation of pleural fluid survivin and XIAP for the diagnosis of malignant pleural effusion. Tumor Biol 2012; 33: Tang Y, Xu L. Superiority and clinical significance of LunX mrna in the diagnosis of malignant pleural effusion caused by pulmonary carcinoma. J Exp Clin Cancer Res 2013; 32: Riantawan P, Sangsayan P, Bangpattanasiri K, Rojanaraweewong P. Limited additive value of pleural fluid carcinoembryonic antigen levels in malignant pleural effusion. Respiration 2000; 67: Yamada N, Kitamoto S, Yokoyama S, Hamada T, Goto M, Tsutsumida H, et al. Epigenetic regulation of mucin genes in human cancer. Clin Epigenet 2011; 2: Hollingsworth MA, Swanson BJ. Mucins in cancer: protection and control of the cell surface. Nat Rev Cancer 2004; 4: Kufe DW. Mucins in cancer: function, prognosis and therapy. Nat Rev Cancer 2009; 9: Baeuerle PA, Gires O. EpCAM (CD326) finding its role in cancer. Br J Cancer 2007; 96: Went PT, Lugli A, Meier S, Bundi M, Mirlacher M, Sauter G, et al. Frequent EpCam protein expression in human carcinomas. Hum Pathol 2004; 35: Went P, Dirnhofer S, Schopf D, Moch H, Spizzo G. Expression and prognostic significance of EpCAM. J Cancer Mol 2008; 3: Zhu W-F, Li J, Yu L-C, Wu Y, Tang X-P, Hu Y-M, et al. Prognostic value of EpCAM/MUC1 mrna-positive cells in nonsmall cell lung cancer patients. Tumor Biol 2014; 35: Berois N, Varangot M, Aizen B, Bstrugo R, Zarantonelli L, Fernandez P, et al. Molecular detection of cancer cells in bone marrow and peripheral blood of patients with operable breast cancer: comparison of CK19, MUC1 and CEA using RT-PCR. Eur J Cancer 2000; 36: Saintigny P, Coulon S, Kambouchner M, Ricci S, Martinot E, Danel C, et al. Real-time RT-PCR detection of CK19, CK7 and MUC1 mrna for diagnosis of lymph node micrometastases in non small cell lung carcinoma. Int J Cancer 2005; 115: Rao CG, Chianese D, Doyle GV, Miller MC, Russell T, Sanders Jr RA, et al. Expression of epithelial cell adhesion molecule in carcinoma cells present in blood and primary and metastatic tumors. Int J Oncol 2005; 27: Woelfle U, Breit E, Zafrakas K, Ott M, Schubert F, Muller V, et al. Bi-specific immunomagnetic enrichment of micrometastatic tumor cell clusters from bone marrow of cancer patients. J Immunol Methods 2005; 300: Yu C-J, Shew J-Y, Liaw Y-S, Kuo S-H, Luh K-T, Yang P-C. Application of mucine quantitative competitive reverse transcription polymerase chain reaction in assisting the diagnosis of malignant pleural effusion. Am J Respir Crit Care Med 2001; 164: Passebosc-Faure K, Li G, Lambert C, Cottier M, Gentil- Perret A, Fournel P, et al. Evaluation of a panel of molecular markers for the diagnosis of malignant serous effusion. Clin Cancer Res 2005; 11: Livak KJ, Schimittgen TD. Analysis of reactive gene expression data using real-time quantitative PCR and 2 -CT method. Method 2001; 25:402 8.