- Email: [email protected]
Diagnostic value of CA 549 in pleural fluid. Comparison with CEA, CA 15.3 and CA 72.4
Lung Cancer 40 (2003) 289 /294 www.elsevier.com/locate/lungcan
Diagnostic value of CA 549 in pleural fluid. Comparison with CEA, CA 15.3 and CA 72.4 Victoria Villena a,*, Angel Lo´pez-Encuentra a, Jose´ Echave-Sustaeta a, Pedro Martı´n-Escribano a, Blanca Ortun˜o-de-Solo b, Juana Estenoz-Alfaro b a
Respiratory Medicine Department, Servicio de Neumologı´a, Hospital 12 de Octubre, Avda de Co´rdoba s/n, Carretera de Andalucı´a Km 5.4, 28041 Madrid, Spain b Nuclear Medicine Department, Hospital 12 de Octubre, Avda de Co´rdoba s/n, 28041 Madrid, Spain Received 30 October 2002; received in revised form 10 February 2003; accepted 12 February 2003
Abstract Several tumor markers have been evaluated in pleural fluid, but their clinical role has not been firmly established. The aim of this study is to determine the diagnostic value of carbohydrate antigen 549 (CA 549) levels in pleural fluid, and to compare it with another previously studied tumor markers: carcinoembryonic antigen (CEA), CA 15.3 and CA 72.4. We prospectively studied 252 patients with pleural effusion: 101 malignant (20 mesothelioma) and 151 of several benign diseases. The levels of the tumor markers were measured by immunoradiometric assays (RIA). CA 549 in pleural fluid has an acceptable sensitivity (0.49), with high specificity (0.99). The best combination of tumor markers for differentiating malignant from benign effusions was CA 549/CEA/CA 15.3, with a sensitivity of 0.65, specificity of 0.99 and accuracy of 0.85. The addition of any one tumor marker assay consistently improved the diagnostic value of cytology. In our study, none of the tumor markers was organ-specific. When mesothelioma and hematological malignancy were ruled-out, the combination of CA 549/CEA/CA 15.3, improved the results up to a sensitivity of 0.77, specificity of 1 and accuracy of 0.92. In conclusion, CA 549 assay has an acceptable sensitivity with high specificity. The best combination of tumor markers in this series with a high relative frequency of mesothelioma and low frequency of breast carcinoma was CA 549/CEA/CA 15.3. Individual tumor markers or their combination increased the sensitivity of pleural cytology. # 2003 Elsevier Science Ireland Ltd. All rights reserved. Keywords: CA 549; CEA; CA 15.3; CA 72.4; Tumor markers; Malignant pleural effusion
1. Introduction Pleural effusion often remains a diagnostic problem after the biochemical and cytological analysis of the pleural fluid. In most series, malignancy is the most common cause of pleural exudates. In order to establish a diagnosis for these patients, several tumor markers have been studied in the last few years. Some of them, such as carcinoembryonic antigen (CEA), CA 15.3, CA
This study has been partially presented at the SEPAR 2001 meeting (A Corun˜a, Spain) and at the European Respiratory Society Annual Congress 2002 (Stockholm). No portion of the manuscript has been published previously. * Corresponding author. Tel.: /34-91-390-8456; fax: /34-91-4695775. E-mail address: [email protected] (V. Villena).
72.4, CA 19.9, mucin-like carcinoma-associated antigen (MCA) or CYFRA 21-1 [1 /6] have a high specificity with cut-off points that offer a moderate sensitivity. Carbohydrate antigen 549 (CA 549) is a high molecular weight acid glycoprotein recognized by two different monoclonal antibodies: BC4E, directed against a human breast cancer line; and BC4N 154, against the membranes of milk fat globules. Serum levels of CA 549 have been widely studied in relation to mammary carcinoma, both at presentation and in detecting relapses [7,8]. Nevertheless, knowledge about the value of CA 549 measurement in pleural fluid is still limited [9]. The aim of this study is to determine the diagnostic value of CA 549 levels in pleural fluid for malignant effusions, and to compare it with another previously studied tumor markers: CEA, CA 15.3 and CA 72.4.
0169-5002/03/$ - see front matter # 2003 Elsevier Science Ireland Ltd. All rights reserved. doi:10.1016/S0169-5002(03)00120-X
290
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
2. Material and methods 2.1. Patients We prospectively included those patients with pleural effusion assessed in our Service from February 1998 to March 2001 in whom at least one diagnostic thoracentesis with measurement of CA 549, CEA, CA 15.3 and CA 72.4 pleural fluid levels was performed. Patients were assessed according to the same diagnostic algorithm [10], which included pleural biopsy, thoracoscopy or thoracotomy for patients with pleural exudates of unknown etiology. A chest, abdominal and pelvic computed tomography, and a bronchoscopy were usually performed to these patients. In addition, chemotherapy responders tumors were looking for by performing a gynecological or testicular physical examination, a mammography (if female), and a prostatic specific antigen (if male). In order to a better knowledge of these tumor markers, we have preferred to classified an origin as ‘‘unknown’’ if the origin of the tumor was likely but not certain. Periodic clinical and radiological follow-ups were performed for patients with effusions of undetermined causes. The clinical physicians did not know CA 549 pleural fluid levels and the laboratory physicians did not know the clinical diagnosis. To be included into the analyzed group, patients were classified into different diagnostic groups based on the following explicit criteria: 1) Neoplastic pleural effusion: Positive cytologic or histologic diagnosis of neoplasm of the pleural space. 2) Tuberculous pleural effusion: Positive Lo¨wenstein culture of pleural fluid or tissue, and/or granulomas on pleural biopsy in the absence of other pleural granulomatous disease and/or high interferon gamma levels in pleural fluid [11] with scarce pleural fluid which precluded pleural biopsy, a favorable clinical course with tuberculous treatment and either extrapleural tuberculosis confirmed by Lo¨wenstein culture or a clinical picture suggestive or tuberculosis. 3) Parapneumonic effusion: Presence of cough, fever and a radiographic pulmonary infiltrate that disappeared with antibiotic treatment. Patients with pleural empyema were also included in this group. 4) Transudate: Standard clinical and biochemical criteria [12]. 5) Miscellaneous: Standard criteria for the diagnosis of other effusions [12]. 6) Pleural effusion of non-neoplastic unknown origin: Pleural effusion of no known cause in patients with either of these criteria: / Non-specific pleuritis studied by thoracoscopy, thoracotomy, or necropsy.
/ No symptoms or recurrence of effusion in a clinical and radiological 1-year follow-up. Patients not included in any of the diagnostic groups, were excluded (n /37). The ethics committee of the 12 de Octubre Hospital approved the study. A informed consent was obtained from patients. 2.2. Tumor markers assay Pleural fluid samples were collected in dry tubes. The samples were immediately taken to the laboratory, centrifuged for 10 min at 1500 /g, and stored frozen at /20 8C until assayed. All markers were assayed using commercial immunoradiometric (RIA) kits: CA 549 (BresmarqTM, Tandem† -R CA 549, USA), CA 72.4 (IRMA-mat CA 72-4, BYK, Sangtec, Germany), CEA (RIA-gnost CEA, Behring, Germany), CA 15.3 (ELSA-CA 15-3, Cis, France). The samples were assayed in duplicate, with dilutions as appropriate and blinded of clinical data. 2.3. Statistical analysis Statistical analysis was performed with SPSS statistical software. The variables did not have a normal distribution according to the Kolmogorov/Smirnov’s test. Median values were compared by the Mann/Whitney U -test. Receiver operating characteristic (ROC) curves were calculated in order to select the cut-off level. The cut-off point for each tumor marker was selected: (a) on the basis of best accuracy; if several cut-off points had the same accuracy, the value with the best specificity was chosen; and (b) to obtain a specificity of 1. In the combinations of tumor markers, the test was considered positive if the cut-off point for any of the markers was exceeded.
3. Results The four tumor markers were measured in pleural fluid from 252 patients, of these 164 were males and 88 females. The mean age was 61 years (range 15 /102, and standard deviation 17.6). Twenty-one percent of patients in the benign group, and 4% of patients with malignant effusions were younger than 40. The etiology of the effusions is shown in Table 1. Neoplastic effusions were 40% of them. The histological types were: 38 adenocarcinomas, 20 mesotheliomas, eight large cell carcinomas, five squamous cell carcinomas, two small cell carcinomas, 18 non-differentiated carcinoma, six non-Hodgkin lymphomas, one Hodgkin’s lymphoma, one chronic myeloid leukemia, one Kaposi’s sarcoma and one unknown. Among the bronchial neoplasms
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
291
Table 1 Etiology of the effusion from the 252 patients included in the study Neoplasms Lung Mesothelioma Breast Ovary Lymphoma or leukemia Rectum Kidney Uterus Bladder Pharynx Kaposi’s sarcoma Unknown Parapneumonic Transudates Tuberculosis Miscellaneous Cardiovascular procedures or diseases Pulmonary embolism Postabdominal surgery Posttrauma Rheumatoid arthritis Uremia Asbestos pleural effusions Pleural amyloidosis Bronchiolitis obliterans organizing pneumonia Churg-Strauss granulomatosis Pleural fibrous localized tumor Pseudochylothorax Trapped lung Several benign causes Non-neoplastic unknown etiology
101 25 20 10 4 8 1 1 1 1 1 1 28 43 38 27 26 6 3 3 3 1 1 1 1 1 1 1 1 1 2 17
there were nine adenocarcinomas, five large cell carcinomas, three squamous cell carcinomas, two small cell carcinomas, and six non-differentiated carcinoma. The median, minimum and maximum levels for the tumor markers in the neoplastic and non-neoplastic groups, are shown in Table 2. The ROC curve for CA 549, CEA, CA 15.3 and CA 72.4 are shown in Fig. 1. The areas under the ROC curves of the four tumor markers were: CA 549: 0.85 (CI 95% 0.80 /0.90), CEA: 0.65 (CI 95% 0.57 /0.72), CA 15.3: 0.79 (CI 95% 0.73 / 0.84), and CA 72.4: 0.73 (CI 95% 0.67 /0.80).
Fig. 1. Receiver operating curves for CA 549, CEA, CA 15.3 and CA 72.4 levels in pleural fluid.
The operating characteristics for the four tumor markers with their cut-off points for achieving the best individual accuracy are shown in Table 3. The patients with benign pleural effusions with tumor marker levels higher than the cut-off points were: one parapneumonic effusion for CA 549 (27.2 IU/ml), one tuberculous effusion for CA 15.3 (52 IU/ml), and two effusions associated to pericardial disease (15.8 and 14.4 IU/ml, respectively) and one in the group of non-neoplastic unknown etiology (12.1 IU/ml) for CA 72.4. The best combination of tumor markers was CA 549/CEA/ CA 15.3. Table 4 shows the sensitivity and accuracy of the four tumor markers with the cut-off points that offered a specificity of 1. Again, the best combination of tumor markers was CA 549/CEA/CA 15.3. Forty-six patients in the neoplastic group had a positive pleural cytology in the first thoracentesis (sensitivity 0.46). The addition of any one tumor marker assay using the cut-off point with higher specificity consistently improved the sensitivity of cytology: CA 549/cytology: 0.63, CEA/cytology: 0.59, CA 15.3/
Table 2 Median, minimum and maximum levels for the tumor markers according to a malignant or benign diagnosis Tumor marker
Diagnosis
Median (IU/ml) (minimum-maximum levels)
P
CA 549
Malignant Benign Malignant Benign Malignant Benign Malignant Benign
20.2 (1.9 /1515) 4.9 (0.1 /27.2) 3.6 (0.1 /22000) 1.9 (0 /30.2) 30.9 (3.6 /1098) 11.1 (0.5 /52) 2.8 (0.1 /4642) 1.4 (0 /15.8)
B/0.001
CEA CA 15.3 CA 72.4
B/0.001 B/0.001 B/0.001
292
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
Table 3 Operating characteristics of each tumor marker and their best combinations Tumor marker
Cut-off point
Sensitivity (CI 95%)
Specificity (CI 95%)
Accuracy (CI 95%)
CA 549 CEA CA 15.3 CA 72.4 CA 549/CEA CEA/CA 15.3 CA 549/CEA/CA 15.3a CA 549/CEA/CA 72.4 CA 549/CEA/CA 15.3/CA 72.4
24 40 45 10
0.49 0.35 0.44 0.35 0.61 0.60 0.65 0.63 0.66
0.99 (0.96 /1) 1 (0.98 /1) 0.99 (0.96 /1) 0.98 (0.94 /1) 0.99 (0.96 /1) 0.99 (0.96 /1) 0.99 (0.95 /1) 0.97 (0.93 /0.99) 0.97 (0.93 /0.99)
0.79 (0.73 /0.83) 0.74 (0.68 /0.79) 0.77 (0.71 /0.82) 0.73 (0.66 /0.78) 0.84 (0.79 /0.88) 0.84 (0.77 /0.88) 0.85a (0.80 /0.89) 0.84 (0.78 /0.88) 0.85 (0.80 /0.89)
IU/ml ng/ml IU/ml IU/ml
(0.38 /0.59) (0.26 /0.45) (0.34 /0.54) (0.26 /0.45) (0.51 /0.71) (0.50 /0.70) (0.55 /0.75) (0.53 /0.73) (0.56 /0.76)
The cut-off points were chosen on the basis of the highest accuracy. a Best combination of tumor markers.
cytology: 0.63, CA 72.4/cytology: 0.57, CA 549/ CEA/cytology: 0.72, CA 549/CEA/CA 15.3/cytology: 0.74. Fifty-five patients from the malignant group had negative first pleural cytologies. Eighteen of these were detected with either CA 549 or CA 15.3, 14 patients with CEA and 12 patients with CA 72.4. Among these 55 patients with negative first pleural cytologies, 31 had at least one of the tumor marker levels above the cut-off level. In 27 of them, the malignant diagnosis was established by either pleural biopsy or repeating the thoracentesis; in two patients by thoracoscopy, in one of them by thoracotomy and in another one by necropsy. Table 5 shows the origin of the tumors with tumor markers above the cut-off point. In our study, none of the tumor markers was organ-specific. After ruling-out the patients with mesothelioma and hematological malignancies, the operating characteristics of the tumor markers and their combinations with a specificity of 1 are shown in Table 6.
4. Discussion Establishing the diagnosis of patients with pleural effusion continues to be a clinical problem. The most common causes of pleural exudates in most series are
malignancy and tuberculosis. In the last few years, several papers have studied the role of biochemical determinations or tumor markers, to increase the diagnostic capacity of pleural fluid analysis, in order to adjust the selection of patients for more aggressive procedures, such as pleural biopsy, thoracoscopy or thoracotomy.
Table 5 Origin of the tumors with tumor marker levels in pleural fluid that were higher than the respective cut-off points Tumor origin
CA 549
CEA
CA 15.3
CA 72.4
Lung Mesothelioma Breast Ovary Lymphoma or leukemia Rectum Kidney Uterus Bladder Pharynx Kaposi’s sarcoma Unknown
11 (44%) 7 (35%) 7 (70%) 3 (75%) 1 (13%) 0 0 1 (100%) 1 (100%) 0 0 13 (46%)
13 (52%) 0 2 (20%) 2 (50%) 0 1 (100%) 0 0 0 0 0 17 (61%)
9 (36%) 6 (30%) 5 (50%) 3 (75%) 1 (13%) 0 1 (100%) 1 (100%) 0 0 0 14 (50%)
7 (28%) 0 3 (30%) 3 (75%) 0 0 0 1 (100%) 1 (100%) 0 0 15 (54%)
Table 6 Table 4 Operating characteristics of each tumor marker and their best combinations
Operating characteristics of each tumor marker and their best combinations for pleural effusions other than mesothelioma and hematological malignancies
Tumor marker
Cut-off point
Sensitivity
Accuracy
Tumor marker
Cut-off point
Sensitivity
Accuracy
CA 549 CEA CA 15.3 CA 72.4 CA 549/CEA CEA/CA 15.3 CA 549/CEA/CA 15.3a CA 549/CEA/CA 72.4 CA 549/CEA/CA 15.3/CA 72.4
28 40 53 16
0.44 0.35 0.40 0.30 0.59 0.56 0.64 0.61 0.64
0.77 0.74 0.76 0.72 0.84 0.83 0.86a 0.85 0.86
CA 549 CEA CA 15.3 CA 72.4 CA 549/CEA CEA/CA 15.3 CA 549/CEA/CA 15.3a CA 549/CEA/CA 72.4 CA 549/CEA/CA 15.3/CA 72.4
28 40 53 16
0.49 0.48 0.45 0.41 0.71 0.68 0.77 0.74 0.77
0.83 0.83 0.82 0.81 0.91 0.90 0.92a 0.92 0.92
IU/ml ng/ml IU/ml IU/ml
The cut-off points were chosen in order to obtain specificity of 1. a Best combination of tumor markers.
IU/ml ng/ml IU/ml IU/ml
The cut-off points were chosen in order to obtain specificity of 1. a Best combination of tumor markers.
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
In our opinion, the differences in the results of tumor marker operating characteristics among various series depend, on several factors. First, is important to consider the laboratory method, which could influence the best cut-off level and the results of the test itself. Second, the results may be influenced by the method of choosing the cut-off point. The ROC curve [2,3,13,14], best specificity [4], 95% specificity [5], 95% percentile [6], or the same level as plasma [15] have been used to select the cut-off level. We believe that malignant diagnoses imply several important clinical implications in treatment, prognosis and/or general management of the patients. This consideration moved us to study two cut-off points: the first one achieved the best accuracy (by ROC curve) for every tumor marker, and the second one was selected to offer a specificity of 1, both for individual tests or for combinations, improving the accuracy of the combinations of the tumor markers. Third, the etiology of the effusions included in the study should be considered. We think not only the variety and frequency of benign effusions but also the origin and histologic type of the malignant effusions are important. In the present series the more common origin were the lung, the breast, lymphomas or the ovarian. According to other series, gastrointestinal tumors were uncommon [16]. In this study, CA 549 had the best accuracy of the four tumor markers studied. We have chosen CEA, CA 72.4 and CA 15.3 to compare with, because these tumor markers had the best diagnostic value in a previous series of our group (2). The sensitivities of CEA, CA 72.4 and CA 15.3 were slightly worse than the sensitivities previously reported by our group [2], or by other authors [3 /5]. Since our method has not been changed, we think these differences could be a consequence of a slight difference in tumor origins in the sample of this study, with a higher frequency of mesotheliomas and hematological malignancies, and a lower frequency of gastrointestinal tumors. Gastrointestinal tumors commonly increase CEA or CA 72.4, while mesotheliomas have usually low levels of both CEA and CA 72.4. In previous reports [2,3,5,9,17 /19], the combination of several tumor markers usually obtained better accuracy than did any one alone. We also analyzed the diagnostic value of these combinations, with two different cut-off points. In both instances, the combination of CA 549, CEA and CA 15.3 offered better results than individual tumor markers or other combinations. The first cut-off point could be used if only one tumor marker is going to be measured, and the second one offers the best accuracy in combinations of several tumor markers. Although performing several tumor markers increases the cost in comparison to cytology of pleural fluid, it is cheaper than thoracoscopy. Their clinical usefulness depends on several points, including the patient age, general status, etc. They could be useful
293
in order to indicate a thoracoscopy in young patients with a tumor marker above the cut-off point, or to assume a high probable malignant diagnosis in a patient with a deteriorated general status and a positive tumor marker. In addition, each tumor marker or their combinations improved the sensitivity of pleural fluid cytology. This fact has been invariably found in other studies [1 / 5,9,14,19,20]. In the group of patients with negative pleural cytologies the determination of tumor markers is even more relevant with clinically significant usefulness modifying the probability of malignancy. Although the frequency of increased levels of each tumor marker depends on the origin of the tumor, none of these tumor markers have been demonstrated to be organ-specific. The frequency of mesotheliomas in our series has been relatively high (20% of neoplasms). Both CA 15.3 and CA 549 are likely to be increased in mesotheliomas. The association of high levels of CA 15.3 with this tumor was also found in a previous study by our group [2] and in the study reported by Mie´douge´ et al. [5] or Alatas et al. [21]. Since experience with CA 549 is much more limited, increased levels of this tumor marker in mesotheliomas have not been previously reported. No mesothelioma increased CEA or CA 72.4 levels above the cut-off points. Hematological malignancies were also highly frequent in our series. Only one patient had high levels of both CA 549 and CA 15.3 and none showed any increase in the other markers. These factors involve not only the operating characteristics of individual tumor markers, but also to those of the combinations. Since the results of the study could be implemented in a setting with few or without any mesotheliomas or hematological malignancies, we also assessed the operating characteristics after ruling-out these two types of tumors (Table 6). The sensitivity of the markers improved remarkably in this situation, but the best combination was similar to that obtained for the complete series. We believe that the frequency of the origins of the tumors in the different settings could modify the best tumor markers combinations. It is likely that series with a relatively high frequency of breast carcinoma or mesothelioma include in the best combinations tumor markers such as CA 549, CA 15.3 or MCA [1,2,5,6], while in series with higher numbers of gastrointestinal tumors include markers such as CA 72.4 [2,15,18]. In conclusion, CA 549 assay has an acceptable sensitivity with high specificity. Individual tumor markers or their combination increase the sensitivity of pleural cytology. The best combination of tumor markers in this series with a high relative frequency of mesothelioma and hematological malignancies has been CA 549/CEA/CA 15.3.
294
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
Acknowledgements This study was supported in part by grants from FEPAR 92/0381.
[11]
References
[12]
[1] Ammon A, Eiffert H, Reil S, Beyer JH, Droese M, Hiddemann W. Tumor-associated antigens in effusions of malignant and benign origin. Clin Invest 1993;71:437 /44. [2] Villena V, Lo´pez Encuentra A, Echave-Sustaeta J, Martı´n Escribano P, Ortun˜o de Solo B, Estenoz Alfaro J. Diagnostic value of CA 72.4, carcinoembryonic antigen, CA 15.3 and CA 19.9 assay in pleural fluid: a study of 207 patients. Cancer 1996;78:736 /40. [3] Romero S, Ferna´ndez C, Arrierio JM, et al. CEA, CA 15-3 and CYFRA 21-1 in serum and pleural fluid of patients with pleural effusions. Eur Respir J 1996;5:17 /23. [4] Ferrer J, Villarino MA, Encabo G, et al. Diagnostic utility of CYFRA 21-1, carcinoembryonic antigen, CA 125, neuron specific enolase, and squamous cell antigen level determinations in the serum and pleural fluid of patients with pleural effusions. Cancer 1999;86:1488 /95. [5] Mie´douge´ M, Rouzaud P, Salama G, et al. Evaluation of seven tumor markers in pleural fluid for the diagnosis of malignant effusions. Br J Cancer 1999;81:1059 /65. [6] Kuralay F, Tokgo¨z Z, Co¨mlekci A. Diagnostic usefulness of tumour marker levels in pleural effusions of malignant and benign origin. Clin Chim Acta 2000;300:43 /5. [7] Pavesi F, Lotzniker M, Scarabelli M, et al. Circulating CA 549 and other associated antigens in breast cancer patients. Oncology 1994;51:18 /21. [8] Martoni A, Zamagni C, Bellanova B, et al. CEA, MCA, CA 15.3 and CA 549 and their combinations in expressing and monitoring metastatic breast cancer: a prospective comparative study. Eur J Cancer 1995;31A:1615 /21. [9] Herna´ndez L, Espasa A, Ferna´ndez C, Candela A, Martı´n C, Romero S. CEA and CA 549 in serum and pleural fluid of patients with pleural effusion. Lung Cancer 2002;36:83 /9. ´ lvarez [10] Villena V, Lo´pez Encuentra A, Echave-Sustaeta J, .A Martı´nez C, Martı´n Escribano P. Estudio prospectivo de 1000
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
pacientes consecutivos con derrame pleural. Etiologı´a del derrame y caracterı´sticas de los pacientes. Arch Bronconeumol 2002;38:21 /6. Villena V, Lo´pez Encuentra A, Echave-Sustaeta J, et al. Interferon gamma in 388 immunocompromised and immunocompetent patients for diagnosing pleural tuberculosis. Eur Respir J 1996;9:2635 /9. Light RW. Clinical manifestations and useful test. In: Light RW, editor. Pleural diseases, 4th ed. Philadelphia: Lippincott Williams and Wilkins, 2001:42 /86. San Jose´ ME, Alvarez D, Valdes L, Sarandeses A, Valle JM, Penela P. Utility of tumour markers in the diagnosis of neoplastic pleural effusion. Clin Chim Acta 1997;265:193 /205. Riantawan P, Sangsayan P, Bangpattanasiri K, Rojanaraweewong P. Limited additive value of pleural fluid carcinoembryonic antigen level in malignant pleural effusion. Respiration 2000;67:24 /9. Cascinu S, Del Ferro E, Barbanti I, Ligi M, Fedeli A, Catalano G. Tumor markers in the diagnosis of malignant serous effusions. Am J Clin Oncol 1997;20:247 /50. Light RW. Pleural effusion related to metastatic malignancies. In: Light RW, editor. Pleural diseases, 4th ed. Philadelphia: Lippincott Williams and Wilkins, 2001:108 /34. Shimokata K, Totani Y, Nakanishi K, Yamamoto M, Hasegawa Y, Kawatsu H, et al. Diagnostic value of cancer antigen 15-3 (CA 15-3) detected by monoclonal antibodies (115D8 and DF3) in exudative pleural effusions. Eur Respir J 1988;1:341 /4. Ferroni P, Szpak C, Greiner JW, Simpson JF, Guadagni F, Johnston WW, et al. CA 72-4 radioimmunoassay in the diagnosis of malignant effusions. Comparison of various tumor markers. Int J Cancer 1990;46:445 /51. Menard O, Dosset B, Jacob C, Martinet Y. Improvement of the diagnosis of the cause of pleural effusion in patients with lung cancer by simultaneous quantification of carcinoembryonic antigen (CEA) and neuron-specific enolase (NSE) pleural levels. Eur J Cancer 1993;29:1806 /9. Mezger J, Permanetter W, Gerbes AL, Wilmanns W, Lamerz R. Tumor associated antigens in diagnosis of serous effusions. J Clin Pathol 1988;41:633 /43. ¨ , Harmanci E, Demir S. ¨ , Metintas M, .C Alatas F, Alatas O ¸ olak O Diagnostic value of CEA, CA 15-3, CA 19-9, CYFRA 21-1, NSE and TSA assay in pleural effusions. Lung Cancer 2001;31:9 /16.
Diagnostic value of CA 549 in pleural fluid. Comparison with CEA, CA 15.3 and CA 72.4 Victoria Villena a,*, Angel Lo´pez-Encuentra a, Jose´ Echave-Sustaeta a, Pedro Martı´n-Escribano a, Blanca Ortun˜o-de-Solo b, Juana Estenoz-Alfaro b a
Respiratory Medicine Department, Servicio de Neumologı´a, Hospital 12 de Octubre, Avda de Co´rdoba s/n, Carretera de Andalucı´a Km 5.4, 28041 Madrid, Spain b Nuclear Medicine Department, Hospital 12 de Octubre, Avda de Co´rdoba s/n, 28041 Madrid, Spain Received 30 October 2002; received in revised form 10 February 2003; accepted 12 February 2003
Abstract Several tumor markers have been evaluated in pleural fluid, but their clinical role has not been firmly established. The aim of this study is to determine the diagnostic value of carbohydrate antigen 549 (CA 549) levels in pleural fluid, and to compare it with another previously studied tumor markers: carcinoembryonic antigen (CEA), CA 15.3 and CA 72.4. We prospectively studied 252 patients with pleural effusion: 101 malignant (20 mesothelioma) and 151 of several benign diseases. The levels of the tumor markers were measured by immunoradiometric assays (RIA). CA 549 in pleural fluid has an acceptable sensitivity (0.49), with high specificity (0.99). The best combination of tumor markers for differentiating malignant from benign effusions was CA 549/CEA/CA 15.3, with a sensitivity of 0.65, specificity of 0.99 and accuracy of 0.85. The addition of any one tumor marker assay consistently improved the diagnostic value of cytology. In our study, none of the tumor markers was organ-specific. When mesothelioma and hematological malignancy were ruled-out, the combination of CA 549/CEA/CA 15.3, improved the results up to a sensitivity of 0.77, specificity of 1 and accuracy of 0.92. In conclusion, CA 549 assay has an acceptable sensitivity with high specificity. The best combination of tumor markers in this series with a high relative frequency of mesothelioma and low frequency of breast carcinoma was CA 549/CEA/CA 15.3. Individual tumor markers or their combination increased the sensitivity of pleural cytology. # 2003 Elsevier Science Ireland Ltd. All rights reserved. Keywords: CA 549; CEA; CA 15.3; CA 72.4; Tumor markers; Malignant pleural effusion
1. Introduction Pleural effusion often remains a diagnostic problem after the biochemical and cytological analysis of the pleural fluid. In most series, malignancy is the most common cause of pleural exudates. In order to establish a diagnosis for these patients, several tumor markers have been studied in the last few years. Some of them, such as carcinoembryonic antigen (CEA), CA 15.3, CA
This study has been partially presented at the SEPAR 2001 meeting (A Corun˜a, Spain) and at the European Respiratory Society Annual Congress 2002 (Stockholm). No portion of the manuscript has been published previously. * Corresponding author. Tel.: /34-91-390-8456; fax: /34-91-4695775. E-mail address: [email protected] (V. Villena).
72.4, CA 19.9, mucin-like carcinoma-associated antigen (MCA) or CYFRA 21-1 [1 /6] have a high specificity with cut-off points that offer a moderate sensitivity. Carbohydrate antigen 549 (CA 549) is a high molecular weight acid glycoprotein recognized by two different monoclonal antibodies: BC4E, directed against a human breast cancer line; and BC4N 154, against the membranes of milk fat globules. Serum levels of CA 549 have been widely studied in relation to mammary carcinoma, both at presentation and in detecting relapses [7,8]. Nevertheless, knowledge about the value of CA 549 measurement in pleural fluid is still limited [9]. The aim of this study is to determine the diagnostic value of CA 549 levels in pleural fluid for malignant effusions, and to compare it with another previously studied tumor markers: CEA, CA 15.3 and CA 72.4.
0169-5002/03/$ - see front matter # 2003 Elsevier Science Ireland Ltd. All rights reserved. doi:10.1016/S0169-5002(03)00120-X
290
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
2. Material and methods 2.1. Patients We prospectively included those patients with pleural effusion assessed in our Service from February 1998 to March 2001 in whom at least one diagnostic thoracentesis with measurement of CA 549, CEA, CA 15.3 and CA 72.4 pleural fluid levels was performed. Patients were assessed according to the same diagnostic algorithm [10], which included pleural biopsy, thoracoscopy or thoracotomy for patients with pleural exudates of unknown etiology. A chest, abdominal and pelvic computed tomography, and a bronchoscopy were usually performed to these patients. In addition, chemotherapy responders tumors were looking for by performing a gynecological or testicular physical examination, a mammography (if female), and a prostatic specific antigen (if male). In order to a better knowledge of these tumor markers, we have preferred to classified an origin as ‘‘unknown’’ if the origin of the tumor was likely but not certain. Periodic clinical and radiological follow-ups were performed for patients with effusions of undetermined causes. The clinical physicians did not know CA 549 pleural fluid levels and the laboratory physicians did not know the clinical diagnosis. To be included into the analyzed group, patients were classified into different diagnostic groups based on the following explicit criteria: 1) Neoplastic pleural effusion: Positive cytologic or histologic diagnosis of neoplasm of the pleural space. 2) Tuberculous pleural effusion: Positive Lo¨wenstein culture of pleural fluid or tissue, and/or granulomas on pleural biopsy in the absence of other pleural granulomatous disease and/or high interferon gamma levels in pleural fluid [11] with scarce pleural fluid which precluded pleural biopsy, a favorable clinical course with tuberculous treatment and either extrapleural tuberculosis confirmed by Lo¨wenstein culture or a clinical picture suggestive or tuberculosis. 3) Parapneumonic effusion: Presence of cough, fever and a radiographic pulmonary infiltrate that disappeared with antibiotic treatment. Patients with pleural empyema were also included in this group. 4) Transudate: Standard clinical and biochemical criteria [12]. 5) Miscellaneous: Standard criteria for the diagnosis of other effusions [12]. 6) Pleural effusion of non-neoplastic unknown origin: Pleural effusion of no known cause in patients with either of these criteria: / Non-specific pleuritis studied by thoracoscopy, thoracotomy, or necropsy.
/ No symptoms or recurrence of effusion in a clinical and radiological 1-year follow-up. Patients not included in any of the diagnostic groups, were excluded (n /37). The ethics committee of the 12 de Octubre Hospital approved the study. A informed consent was obtained from patients. 2.2. Tumor markers assay Pleural fluid samples were collected in dry tubes. The samples were immediately taken to the laboratory, centrifuged for 10 min at 1500 /g, and stored frozen at /20 8C until assayed. All markers were assayed using commercial immunoradiometric (RIA) kits: CA 549 (BresmarqTM, Tandem† -R CA 549, USA), CA 72.4 (IRMA-mat CA 72-4, BYK, Sangtec, Germany), CEA (RIA-gnost CEA, Behring, Germany), CA 15.3 (ELSA-CA 15-3, Cis, France). The samples were assayed in duplicate, with dilutions as appropriate and blinded of clinical data. 2.3. Statistical analysis Statistical analysis was performed with SPSS statistical software. The variables did not have a normal distribution according to the Kolmogorov/Smirnov’s test. Median values were compared by the Mann/Whitney U -test. Receiver operating characteristic (ROC) curves were calculated in order to select the cut-off level. The cut-off point for each tumor marker was selected: (a) on the basis of best accuracy; if several cut-off points had the same accuracy, the value with the best specificity was chosen; and (b) to obtain a specificity of 1. In the combinations of tumor markers, the test was considered positive if the cut-off point for any of the markers was exceeded.
3. Results The four tumor markers were measured in pleural fluid from 252 patients, of these 164 were males and 88 females. The mean age was 61 years (range 15 /102, and standard deviation 17.6). Twenty-one percent of patients in the benign group, and 4% of patients with malignant effusions were younger than 40. The etiology of the effusions is shown in Table 1. Neoplastic effusions were 40% of them. The histological types were: 38 adenocarcinomas, 20 mesotheliomas, eight large cell carcinomas, five squamous cell carcinomas, two small cell carcinomas, 18 non-differentiated carcinoma, six non-Hodgkin lymphomas, one Hodgkin’s lymphoma, one chronic myeloid leukemia, one Kaposi’s sarcoma and one unknown. Among the bronchial neoplasms
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
291
Table 1 Etiology of the effusion from the 252 patients included in the study Neoplasms Lung Mesothelioma Breast Ovary Lymphoma or leukemia Rectum Kidney Uterus Bladder Pharynx Kaposi’s sarcoma Unknown Parapneumonic Transudates Tuberculosis Miscellaneous Cardiovascular procedures or diseases Pulmonary embolism Postabdominal surgery Posttrauma Rheumatoid arthritis Uremia Asbestos pleural effusions Pleural amyloidosis Bronchiolitis obliterans organizing pneumonia Churg-Strauss granulomatosis Pleural fibrous localized tumor Pseudochylothorax Trapped lung Several benign causes Non-neoplastic unknown etiology
101 25 20 10 4 8 1 1 1 1 1 1 28 43 38 27 26 6 3 3 3 1 1 1 1 1 1 1 1 1 2 17
there were nine adenocarcinomas, five large cell carcinomas, three squamous cell carcinomas, two small cell carcinomas, and six non-differentiated carcinoma. The median, minimum and maximum levels for the tumor markers in the neoplastic and non-neoplastic groups, are shown in Table 2. The ROC curve for CA 549, CEA, CA 15.3 and CA 72.4 are shown in Fig. 1. The areas under the ROC curves of the four tumor markers were: CA 549: 0.85 (CI 95% 0.80 /0.90), CEA: 0.65 (CI 95% 0.57 /0.72), CA 15.3: 0.79 (CI 95% 0.73 / 0.84), and CA 72.4: 0.73 (CI 95% 0.67 /0.80).
Fig. 1. Receiver operating curves for CA 549, CEA, CA 15.3 and CA 72.4 levels in pleural fluid.
The operating characteristics for the four tumor markers with their cut-off points for achieving the best individual accuracy are shown in Table 3. The patients with benign pleural effusions with tumor marker levels higher than the cut-off points were: one parapneumonic effusion for CA 549 (27.2 IU/ml), one tuberculous effusion for CA 15.3 (52 IU/ml), and two effusions associated to pericardial disease (15.8 and 14.4 IU/ml, respectively) and one in the group of non-neoplastic unknown etiology (12.1 IU/ml) for CA 72.4. The best combination of tumor markers was CA 549/CEA/ CA 15.3. Table 4 shows the sensitivity and accuracy of the four tumor markers with the cut-off points that offered a specificity of 1. Again, the best combination of tumor markers was CA 549/CEA/CA 15.3. Forty-six patients in the neoplastic group had a positive pleural cytology in the first thoracentesis (sensitivity 0.46). The addition of any one tumor marker assay using the cut-off point with higher specificity consistently improved the sensitivity of cytology: CA 549/cytology: 0.63, CEA/cytology: 0.59, CA 15.3/
Table 2 Median, minimum and maximum levels for the tumor markers according to a malignant or benign diagnosis Tumor marker
Diagnosis
Median (IU/ml) (minimum-maximum levels)
P
CA 549
Malignant Benign Malignant Benign Malignant Benign Malignant Benign
20.2 (1.9 /1515) 4.9 (0.1 /27.2) 3.6 (0.1 /22000) 1.9 (0 /30.2) 30.9 (3.6 /1098) 11.1 (0.5 /52) 2.8 (0.1 /4642) 1.4 (0 /15.8)
B/0.001
CEA CA 15.3 CA 72.4
B/0.001 B/0.001 B/0.001
292
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
Table 3 Operating characteristics of each tumor marker and their best combinations Tumor marker
Cut-off point
Sensitivity (CI 95%)
Specificity (CI 95%)
Accuracy (CI 95%)
CA 549 CEA CA 15.3 CA 72.4 CA 549/CEA CEA/CA 15.3 CA 549/CEA/CA 15.3a CA 549/CEA/CA 72.4 CA 549/CEA/CA 15.3/CA 72.4
24 40 45 10
0.49 0.35 0.44 0.35 0.61 0.60 0.65 0.63 0.66
0.99 (0.96 /1) 1 (0.98 /1) 0.99 (0.96 /1) 0.98 (0.94 /1) 0.99 (0.96 /1) 0.99 (0.96 /1) 0.99 (0.95 /1) 0.97 (0.93 /0.99) 0.97 (0.93 /0.99)
0.79 (0.73 /0.83) 0.74 (0.68 /0.79) 0.77 (0.71 /0.82) 0.73 (0.66 /0.78) 0.84 (0.79 /0.88) 0.84 (0.77 /0.88) 0.85a (0.80 /0.89) 0.84 (0.78 /0.88) 0.85 (0.80 /0.89)
IU/ml ng/ml IU/ml IU/ml
(0.38 /0.59) (0.26 /0.45) (0.34 /0.54) (0.26 /0.45) (0.51 /0.71) (0.50 /0.70) (0.55 /0.75) (0.53 /0.73) (0.56 /0.76)
The cut-off points were chosen on the basis of the highest accuracy. a Best combination of tumor markers.
cytology: 0.63, CA 72.4/cytology: 0.57, CA 549/ CEA/cytology: 0.72, CA 549/CEA/CA 15.3/cytology: 0.74. Fifty-five patients from the malignant group had negative first pleural cytologies. Eighteen of these were detected with either CA 549 or CA 15.3, 14 patients with CEA and 12 patients with CA 72.4. Among these 55 patients with negative first pleural cytologies, 31 had at least one of the tumor marker levels above the cut-off level. In 27 of them, the malignant diagnosis was established by either pleural biopsy or repeating the thoracentesis; in two patients by thoracoscopy, in one of them by thoracotomy and in another one by necropsy. Table 5 shows the origin of the tumors with tumor markers above the cut-off point. In our study, none of the tumor markers was organ-specific. After ruling-out the patients with mesothelioma and hematological malignancies, the operating characteristics of the tumor markers and their combinations with a specificity of 1 are shown in Table 6.
4. Discussion Establishing the diagnosis of patients with pleural effusion continues to be a clinical problem. The most common causes of pleural exudates in most series are
malignancy and tuberculosis. In the last few years, several papers have studied the role of biochemical determinations or tumor markers, to increase the diagnostic capacity of pleural fluid analysis, in order to adjust the selection of patients for more aggressive procedures, such as pleural biopsy, thoracoscopy or thoracotomy.
Table 5 Origin of the tumors with tumor marker levels in pleural fluid that were higher than the respective cut-off points Tumor origin
CA 549
CEA
CA 15.3
CA 72.4
Lung Mesothelioma Breast Ovary Lymphoma or leukemia Rectum Kidney Uterus Bladder Pharynx Kaposi’s sarcoma Unknown
11 (44%) 7 (35%) 7 (70%) 3 (75%) 1 (13%) 0 0 1 (100%) 1 (100%) 0 0 13 (46%)
13 (52%) 0 2 (20%) 2 (50%) 0 1 (100%) 0 0 0 0 0 17 (61%)
9 (36%) 6 (30%) 5 (50%) 3 (75%) 1 (13%) 0 1 (100%) 1 (100%) 0 0 0 14 (50%)
7 (28%) 0 3 (30%) 3 (75%) 0 0 0 1 (100%) 1 (100%) 0 0 15 (54%)
Table 6 Table 4 Operating characteristics of each tumor marker and their best combinations
Operating characteristics of each tumor marker and their best combinations for pleural effusions other than mesothelioma and hematological malignancies
Tumor marker
Cut-off point
Sensitivity
Accuracy
Tumor marker
Cut-off point
Sensitivity
Accuracy
CA 549 CEA CA 15.3 CA 72.4 CA 549/CEA CEA/CA 15.3 CA 549/CEA/CA 15.3a CA 549/CEA/CA 72.4 CA 549/CEA/CA 15.3/CA 72.4
28 40 53 16
0.44 0.35 0.40 0.30 0.59 0.56 0.64 0.61 0.64
0.77 0.74 0.76 0.72 0.84 0.83 0.86a 0.85 0.86
CA 549 CEA CA 15.3 CA 72.4 CA 549/CEA CEA/CA 15.3 CA 549/CEA/CA 15.3a CA 549/CEA/CA 72.4 CA 549/CEA/CA 15.3/CA 72.4
28 40 53 16
0.49 0.48 0.45 0.41 0.71 0.68 0.77 0.74 0.77
0.83 0.83 0.82 0.81 0.91 0.90 0.92a 0.92 0.92
IU/ml ng/ml IU/ml IU/ml
The cut-off points were chosen in order to obtain specificity of 1. a Best combination of tumor markers.
IU/ml ng/ml IU/ml IU/ml
The cut-off points were chosen in order to obtain specificity of 1. a Best combination of tumor markers.
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
In our opinion, the differences in the results of tumor marker operating characteristics among various series depend, on several factors. First, is important to consider the laboratory method, which could influence the best cut-off level and the results of the test itself. Second, the results may be influenced by the method of choosing the cut-off point. The ROC curve [2,3,13,14], best specificity [4], 95% specificity [5], 95% percentile [6], or the same level as plasma [15] have been used to select the cut-off level. We believe that malignant diagnoses imply several important clinical implications in treatment, prognosis and/or general management of the patients. This consideration moved us to study two cut-off points: the first one achieved the best accuracy (by ROC curve) for every tumor marker, and the second one was selected to offer a specificity of 1, both for individual tests or for combinations, improving the accuracy of the combinations of the tumor markers. Third, the etiology of the effusions included in the study should be considered. We think not only the variety and frequency of benign effusions but also the origin and histologic type of the malignant effusions are important. In the present series the more common origin were the lung, the breast, lymphomas or the ovarian. According to other series, gastrointestinal tumors were uncommon [16]. In this study, CA 549 had the best accuracy of the four tumor markers studied. We have chosen CEA, CA 72.4 and CA 15.3 to compare with, because these tumor markers had the best diagnostic value in a previous series of our group (2). The sensitivities of CEA, CA 72.4 and CA 15.3 were slightly worse than the sensitivities previously reported by our group [2], or by other authors [3 /5]. Since our method has not been changed, we think these differences could be a consequence of a slight difference in tumor origins in the sample of this study, with a higher frequency of mesotheliomas and hematological malignancies, and a lower frequency of gastrointestinal tumors. Gastrointestinal tumors commonly increase CEA or CA 72.4, while mesotheliomas have usually low levels of both CEA and CA 72.4. In previous reports [2,3,5,9,17 /19], the combination of several tumor markers usually obtained better accuracy than did any one alone. We also analyzed the diagnostic value of these combinations, with two different cut-off points. In both instances, the combination of CA 549, CEA and CA 15.3 offered better results than individual tumor markers or other combinations. The first cut-off point could be used if only one tumor marker is going to be measured, and the second one offers the best accuracy in combinations of several tumor markers. Although performing several tumor markers increases the cost in comparison to cytology of pleural fluid, it is cheaper than thoracoscopy. Their clinical usefulness depends on several points, including the patient age, general status, etc. They could be useful
293
in order to indicate a thoracoscopy in young patients with a tumor marker above the cut-off point, or to assume a high probable malignant diagnosis in a patient with a deteriorated general status and a positive tumor marker. In addition, each tumor marker or their combinations improved the sensitivity of pleural fluid cytology. This fact has been invariably found in other studies [1 / 5,9,14,19,20]. In the group of patients with negative pleural cytologies the determination of tumor markers is even more relevant with clinically significant usefulness modifying the probability of malignancy. Although the frequency of increased levels of each tumor marker depends on the origin of the tumor, none of these tumor markers have been demonstrated to be organ-specific. The frequency of mesotheliomas in our series has been relatively high (20% of neoplasms). Both CA 15.3 and CA 549 are likely to be increased in mesotheliomas. The association of high levels of CA 15.3 with this tumor was also found in a previous study by our group [2] and in the study reported by Mie´douge´ et al. [5] or Alatas et al. [21]. Since experience with CA 549 is much more limited, increased levels of this tumor marker in mesotheliomas have not been previously reported. No mesothelioma increased CEA or CA 72.4 levels above the cut-off points. Hematological malignancies were also highly frequent in our series. Only one patient had high levels of both CA 549 and CA 15.3 and none showed any increase in the other markers. These factors involve not only the operating characteristics of individual tumor markers, but also to those of the combinations. Since the results of the study could be implemented in a setting with few or without any mesotheliomas or hematological malignancies, we also assessed the operating characteristics after ruling-out these two types of tumors (Table 6). The sensitivity of the markers improved remarkably in this situation, but the best combination was similar to that obtained for the complete series. We believe that the frequency of the origins of the tumors in the different settings could modify the best tumor markers combinations. It is likely that series with a relatively high frequency of breast carcinoma or mesothelioma include in the best combinations tumor markers such as CA 549, CA 15.3 or MCA [1,2,5,6], while in series with higher numbers of gastrointestinal tumors include markers such as CA 72.4 [2,15,18]. In conclusion, CA 549 assay has an acceptable sensitivity with high specificity. Individual tumor markers or their combination increase the sensitivity of pleural cytology. The best combination of tumor markers in this series with a high relative frequency of mesothelioma and hematological malignancies has been CA 549/CEA/CA 15.3.
294
V. Villena et al. / Lung Cancer 40 (2003) 289 /294
Acknowledgements This study was supported in part by grants from FEPAR 92/0381.
[11]
References
[12]
[1] Ammon A, Eiffert H, Reil S, Beyer JH, Droese M, Hiddemann W. Tumor-associated antigens in effusions of malignant and benign origin. Clin Invest 1993;71:437 /44. [2] Villena V, Lo´pez Encuentra A, Echave-Sustaeta J, Martı´n Escribano P, Ortun˜o de Solo B, Estenoz Alfaro J. Diagnostic value of CA 72.4, carcinoembryonic antigen, CA 15.3 and CA 19.9 assay in pleural fluid: a study of 207 patients. Cancer 1996;78:736 /40. [3] Romero S, Ferna´ndez C, Arrierio JM, et al. CEA, CA 15-3 and CYFRA 21-1 in serum and pleural fluid of patients with pleural effusions. Eur Respir J 1996;5:17 /23. [4] Ferrer J, Villarino MA, Encabo G, et al. Diagnostic utility of CYFRA 21-1, carcinoembryonic antigen, CA 125, neuron specific enolase, and squamous cell antigen level determinations in the serum and pleural fluid of patients with pleural effusions. Cancer 1999;86:1488 /95. [5] Mie´douge´ M, Rouzaud P, Salama G, et al. Evaluation of seven tumor markers in pleural fluid for the diagnosis of malignant effusions. Br J Cancer 1999;81:1059 /65. [6] Kuralay F, Tokgo¨z Z, Co¨mlekci A. Diagnostic usefulness of tumour marker levels in pleural effusions of malignant and benign origin. Clin Chim Acta 2000;300:43 /5. [7] Pavesi F, Lotzniker M, Scarabelli M, et al. Circulating CA 549 and other associated antigens in breast cancer patients. Oncology 1994;51:18 /21. [8] Martoni A, Zamagni C, Bellanova B, et al. CEA, MCA, CA 15.3 and CA 549 and their combinations in expressing and monitoring metastatic breast cancer: a prospective comparative study. Eur J Cancer 1995;31A:1615 /21. [9] Herna´ndez L, Espasa A, Ferna´ndez C, Candela A, Martı´n C, Romero S. CEA and CA 549 in serum and pleural fluid of patients with pleural effusion. Lung Cancer 2002;36:83 /9. ´ lvarez [10] Villena V, Lo´pez Encuentra A, Echave-Sustaeta J, .A Martı´nez C, Martı´n Escribano P. Estudio prospectivo de 1000
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
pacientes consecutivos con derrame pleural. Etiologı´a del derrame y caracterı´sticas de los pacientes. Arch Bronconeumol 2002;38:21 /6. Villena V, Lo´pez Encuentra A, Echave-Sustaeta J, et al. Interferon gamma in 388 immunocompromised and immunocompetent patients for diagnosing pleural tuberculosis. Eur Respir J 1996;9:2635 /9. Light RW. Clinical manifestations and useful test. In: Light RW, editor. Pleural diseases, 4th ed. Philadelphia: Lippincott Williams and Wilkins, 2001:42 /86. San Jose´ ME, Alvarez D, Valdes L, Sarandeses A, Valle JM, Penela P. Utility of tumour markers in the diagnosis of neoplastic pleural effusion. Clin Chim Acta 1997;265:193 /205. Riantawan P, Sangsayan P, Bangpattanasiri K, Rojanaraweewong P. Limited additive value of pleural fluid carcinoembryonic antigen level in malignant pleural effusion. Respiration 2000;67:24 /9. Cascinu S, Del Ferro E, Barbanti I, Ligi M, Fedeli A, Catalano G. Tumor markers in the diagnosis of malignant serous effusions. Am J Clin Oncol 1997;20:247 /50. Light RW. Pleural effusion related to metastatic malignancies. In: Light RW, editor. Pleural diseases, 4th ed. Philadelphia: Lippincott Williams and Wilkins, 2001:108 /34. Shimokata K, Totani Y, Nakanishi K, Yamamoto M, Hasegawa Y, Kawatsu H, et al. Diagnostic value of cancer antigen 15-3 (CA 15-3) detected by monoclonal antibodies (115D8 and DF3) in exudative pleural effusions. Eur Respir J 1988;1:341 /4. Ferroni P, Szpak C, Greiner JW, Simpson JF, Guadagni F, Johnston WW, et al. CA 72-4 radioimmunoassay in the diagnosis of malignant effusions. Comparison of various tumor markers. Int J Cancer 1990;46:445 /51. Menard O, Dosset B, Jacob C, Martinet Y. Improvement of the diagnosis of the cause of pleural effusion in patients with lung cancer by simultaneous quantification of carcinoembryonic antigen (CEA) and neuron-specific enolase (NSE) pleural levels. Eur J Cancer 1993;29:1806 /9. Mezger J, Permanetter W, Gerbes AL, Wilmanns W, Lamerz R. Tumor associated antigens in diagnosis of serous effusions. J Clin Pathol 1988;41:633 /43. ¨ , Harmanci E, Demir S. ¨ , Metintas M, .C Alatas F, Alatas O ¸ olak O Diagnostic value of CEA, CA 15-3, CA 19-9, CYFRA 21-1, NSE and TSA assay in pleural effusions. Lung Cancer 2001;31:9 /16.