Malignant pleural effusion (PE) is a frequent. Diagnostic Value of CYFRA 21 1 Tumor Marker and CEA in Pleural Effusion Due to Mesothelioma*

Transcription

1 Diagnostic Value of CYFRA 21 1 Tumor Marker and CEA in Pleural Effusion Due to Mesothelioma* Michela Paganuzzi, PhD; Marina Onetto, MD; Paola Marroni, PhD; Rosa Filiberti, PhD; Elisabetta Tassara, MD; Stefano Parodi, PhD; and Raffaella Felletti, MD Study objective: The aim of our study was to assess the clinical value of CYFRA 21 1 tumor marker and carcinoembryonic antigen (CEA) as diagnostic tools that are complementary to cytology in the diagnosis of malignant mesotheliomas. Patients: We measured CEA and CYFRA 21 1 in the pleural effusions (PEs) and serum of 106 patients (benign lung disease, 34 patients; bronchogenic and metastatic carcinoma, 40 patients; mesothelioma, 32 patients). Methods: CEA and CYFRA 21 1 levels were determined by means of two commercial enzyme immunoassays. Results: The cutoff levels of CYFRA 21 1 and CEA in malignant PEs, selected on the basis of the best diagnostic efficacy, were 41.9 ng/ml and 5.0 ng/ml, respectively. In all neoplastic PEs, CYFRA 21 1 and CEA sensitivity was 78% and 30.6%, respectively, with a specificity of 80% and 91%, respectively. The sensitivity of CYFRA 21 1 and CEA in patients with mesothelioma was 87.5% and 3.1%, respectively. The results of the CYFRA 21 1 assay were positive in 17 of 19 cases of mesothelioma (89.5%) with a negative or uncertain cytology. The association of the tumor marker assay and the cytology allowed a correct diagnosis in 30 of 32 cases of mesothelioma (93.7%). Conclusion: This study suggests that CYFRA 21 1 would provide a useful parameter for the differential diagnosis between benign and malignant PE from mesothelioma when the result of cytology is negative or uncertain and the clinical context does not allow a more aggressive approach. Moreover, the association of CYFRA 21 1 with CEA could provide details for a differential diagnosis between mesotheliomas and carcinomas. In fact, an elevated CYFRA 21 1 level with a low CEA level is highly suggestive of mesothelioma, whereas high levels of CEA alone or high levels of both the markers suggest a diagnosis of malignant PE, excluding mesothelioma. (CHEST 2001; 119: ) Key words: CEA; CYFRA 21 1; mesothelioma; pleural effusion; tumor marker Abbreviations: CEA carcinoembryonic antigen; PE pleural effusion; ROC receiver operating characteristic Malignant pleural effusion (PE) is a frequent occurrence in patients with cancer and may be either a sign of the disease or a complication of an already diagnosed malignancy. 1 Virtually all primary malignancies have been reported to metastasize to the pleura. Some reports suggest that as many as 50% of patients with lung or *From the Clinical Pathology Laboratory (Drs. Paganuzzi, Onetto, and Marroni), the Department of Environmental Epidemiology (Drs. Filiberti and Parodi), and the Thoracic Endoscopy Service (Dr. Tassara), National Institute for Cancer Research, Genoa, Italy; and the Department of Respiratory Disease (Dr. Felletti), A.O. San Martino Hospital, Genoa, Italy. Manuscript received February 10, 2000; revision accepted October 17, Correspondence to: Michela Paganuzzi, PhD, Clinical Pathology Laboratory, National Institute for Cancer Research, Largo R. Benzi 10, Genoa, Italy; patclin@hp380.ist.unige.it breast cancer will develop PEs at some time during the course of their disease. 2 The sine qua non condition for the diagnosis of malignant PE is the cytologic demonstration of malignant cells in pleural fluid; the absence of malignant cells, however, does not exclude malignancy. The difficulties arise in differentiating benign reactive mesothelial cells from malignant mesothelioma cells and metastatic carcinoma cells. 3 Blind pleural biopsy, which is the alternative method in the case of negative cytology results, performs poorly and has a low sensitivity (from 35 to 65%). 4,5 The recent introduction of videothoracoscopy has improved the efficacy of both the diagnosis and treatment of malignant PE. 6 However, it is burdened with the risk for complications such as 1138 Clinical Investigations

2 pneumothorax, pleural shock, hemothorax, and subcutaneous emphysema, as well as possible cancer implantation in 4% of patients. 7 Furthermore, videothoracoscopy with multiple pleural biopsies is an aggressive and expensive method of investigation, and is not available everywhere. The clinical value of serum tumor markers for diagnosis, follow-up, and treatment monitoring of lung cancer patients already has been demonstrated in many studies. 8,9 The measurement of tumor markers in PE represents an effort to improve effectiveness of the differential diagnosis of neoplastic effusions. Several authors have demonstrated the clinical usefulness of measuring different tumor markers in the diagnosis of malignant PE. Those reported series, however, included mostly cases of adenocarcinoma and only a few cases of mesothelioma. The incidence of this neoplasia is relatively low, but it is increasing in certain geographic areas mostly because of an intensive occupational exposure to asbestos. 14 In our region, the incidence of malignant mesothelioma is high, and it is one of the highest values in Europe, with an age-standardized rate of 5.0 per 100,000 men and 0.9 per 100,000 women. 15 In the present study, we aimed to assess the clinical value of CYFRA 21 1 tumor marker and carcinoembryonic antigen (CEA) as diagnostic tools that are complementary to cytologic findings in the diagnosis of malignant mesotheliomas. For this purpose, we measured the tumor markers in the serum and PE of patients with mesothelioma and compared them with those of patients with pleural metastatic carcinoma and benign pathologic conditions. Patients Materials and Methods We measured CEA and CYFRA 21 1 in the serum and PE of 106 consecutive patients admitted for diagnosis to the Department of Respiratory Disease of San Martino Hospital and to the Thoracic Endoscopy Service of the National Institute for Cancer Research of Genoa from 1995 to Informed consent was obtained from all the subjects. Thirty-four patients had benign PE, and 72 patients had malignant PE. Of these latter patients, 32 malignant PEs were from mesothelioma (Table 1). All the PEs were examined through cytologic analysis. When the result of the cytologic examination was negative or in doubt, patients underwent blind pleural biopsies and/or thoracoscopic-guided biopsies. The PEs were considered to be positive according to the presence of malignant cells as determined by cytologic analysis and/or pleural biopsy. The diagnosis of benign disease was confirmed on the basis of instrumental investigations (CT scan), bacteriologic examination, response to therapy, and clinical follow-up. Tumor Marker Assays Table 1 Description of Patients Diagnosis PEs and blood samples were centrifuged, and aliquots of supernatants and sera were stored at 80 C until tested. CYFRA 21 1 and CEA were measured using two commercially available immunoenzymatic kits. CYFRA 21 1 was measured with a semiautomated system (model ES 300; Boehringer Mannheim; Mannheim, Germany) using a two-step sandwich method with the streptavidin-biotin technique. The assay utilized two mouse monoclonal antibodies, KS 19 1 and BM 19 21, that were directed against two different epitopes of a soluble fragment of cytokeratin 19. CEA was measured using a kit (AXSYM CEA kit; Abbott Laboratories; Abbott Park, IL) that is based on the microparticle enzyme immunoassay technology. The assay of each marker was performed according to the protocol recommended by the manufacturer. Results were expressed in nanograms per milliliter. Statistical Analysis Differences in the median values between the groups under study were checked using the Kruskall-Wallis test, while the Mann-Whitney U test was used for comparison of the two groups. Spearman rank correlation coefficients for the tumor markers were calculated separately for the serum and the PE. The threshold for each marker was selected on the basis of the best diagnostic efficacy (ie, on the basis of the more equilibrated combination of sensitivity and specificity). In order to compare the performance of the tumor markers, receiver operating characteristic (ROC) curves were constructed. Results Patients, No. Mesothelioma 32 Pleural metastatic carcinoma 40 Lung adenocarcinoma 13 Small-cell lung carcinoma 1 Squamous cell lung carcinoma 2 Lung tumor* 7 Metastasis from adenocarcinoma 17 Benign lung disease 34 Tuberculosis 6 Pleurisy 19 Cardiac failure 3 Liver cirrhosis 2 Other 4 Total 106 *Lung tumors not histologically defined. Includes patients with metastases of the breast (6), stomach (2), colon (2), pancreas (1), uterus (1), ovary (3), parotid gland (1), and kidney (1). Diseases include posttraumatic injury (1), pulmonary embolism (1), asbestosis (1), and empyema (1). The median concentration and range of CEA and CYFRA 21 1 in the serum and PE of patients with malignant and benign disease are reported in Table 2. The PE CYFRA 21 1 concentration was significantly higher in patients with mesothelioma than in CHEST / 119 / 4/ APRIL,

3 Table 2 CYFRA 21-1 and CEA Levels in PE and Serum in Neoplastic and Benign Diseases* Marker Mesothelioma Lung Carcinoma Pleural Metastasis Benign Disease Serum, ng/ml CEA 1.3 ( ) 2.2 ( ) 2.6 (0.6 3,202) 1.7 ( ) CYFRA (1.2 84) 5.1 ( ) 7.0 ( ) 2.5 ( ) PE, ng/ml CEA 0.8 (0.2 7) 6.3 (0.2 1,465.2) 6.0 (0.4 43,850) 1.2 ( ) CYFRA (10.1 5,180) 75.0 (7.7 3,198) 94.2 (4.1 26,750) 21.7 ( ) *Values are given as the median (minimum to maximum values). those with benign pleurisy (p 0.001) and with carcinoma (p 0.02). In serum, a significant difference was observed between patients with mesotheliomas and benign diseases (p ), but not those with carcinomas. The CEA values both in the serum and in PE were significantly higher in patients with carcinomas than in those with mesotheliomas (p 0.001). The distribution of CYFRA 21 1 and CEA in PEs is shown in Figure 1. The ROC curves analysis showed the performance of PE CYFRA 21 1 and CEA in mesotheliomas (Fig 2). We evaluated the diagnostic efficacies in serum and PE for CYFRA 21 1 and CEA in all malignancies at different cutoff levels. The sensitivity of the two markers was higher in PEs than in sera (not significant) at any cutoff level. The best diagnostic efficacy of CYFRA 21 1 in PEs (78%) was observed at the cutoff value of 41.9 ng/ml, which achieved a sensitivity of 78% and a specificity of 80%. The sensitivity in patients with carcinomas was 67.5%, and in those with mesotheliomas it was 87.5%. For CEA, we chose the diagnostic accuracy (49%) obtained with the conventional cutoff of 5.0 ng/ml. The sensitivity was 30.6% in all malignancies, 52.5% in carcinomas, and 3.1% in mesotheliomas, with a specificity of 91% (Table 3). The performance of CEA and CYFRA 21 1 in PEs was compared with the cytologic analysis. Forty of the 72 patients with neoplastic disease (55.6%) had negative results of cytologic analysis, 26 patients (36.1%) had positive results, and 6 patients (8.3%) had uncertain behavior. The positivity of CEA and CYFRA 21 1 in the 46 patients with negative or uncertain cytologic analysis results is reported in Table 4. Considering mesotheliomas alone, 13 patients (41%) had positive results of cytologic analysis, 16 patients (50%) had negative results, and 3 patients (9%) had uncertain behavior. Seventeen of 19 patients (89.5%) with negative or uncertain results of cytologic analysis showed positive CYFRA 21 1 concentrations. Figure 1. Distribution of CEA and CYFRA 21 1 concentrations in PEs from carcinomas (F), mesotheliomas (Œ), and benign diseases (E). The cutoff values (CEA, 5.0 ng/ml; CYFRA 21 1, 41.9 ng/ml) are denoted by horizontal and vertical lines, respectively. Figure 2. ROC curves of CEA and CYFRA 21 1 in the differential diagnosis between mesotheliomas and benign PEs Clinical Investigations

4 Table 3 Diagnostic Sensitivity of CEA and CYFRA 21-1 in PEs* Disease CEA, % CYFRA 21-1, % All malignancies Carcinomas Lung carcinomas Pleural metastases Mesotheliomas *The cutoff level for CEA is 5 ng/ml, and that for CYFRA 21-1 is 41.9 ng/ml. The specificity at these cutoff values was 91% for CEA and 80% for CYFRA The cytologic examination combined with the measurement of PE CEA and CYFRA 21 1 increased the detection rate of mesotheliomas to 93.7%. Discussion The differential diagnosis between benign and malignant effusions represents a critical clinical problem. Cytologic analysis is the method usually adopted to identify malignant cells in a PE, but it seems not to be sensitive enough (40 to 60%). 3 In the case of uncertainty, blind or thoracoscopic-guided biopsy should be used. This procedure is highly sensitive, but, unfortunately, it is also invasive and limited to specialized centers. The results of previous studies have suggested that a pool of different markers in the PE can improve the diagnostic power of cytologic analysis in detecting malignant PEs CEA was found to be a complementary tool to immunohistochemical studies for differentiating between PEs resulting from metastatic tumors and from malignant mesothelioma. 13,19 The CYFRA 21 1 assay, which detects a soluble fragment of cytokeratin 19 that is expressed by all histologic types of lung cancer, 20 was indicated as the most available tumor marker in PEs from epithelial carcinomas, especially squamous cell carcinomas. 11,21 Overall, our findings agree with those of other Table 4 Cases Showing CEA and/or CYFRA 21-1 Values Higher Than the Cutoff Threshold in the 46 Malignant PEs With Negative or Uncertain Cytology Results* Malignancies Patients, No. CEA CYFRA 21-1 CEA and/or CYFRA 21-1 Lung carcinomas Pleural metastases Mesotheliomas Total *See Table 3 and text for cutoff levels, specificity, and sensitivity. authors. No correlation was found between CEA and CYFRA 21 1 values, thus confirming that the two molecules provide different information on tumor behavior. Both CEA and CYFRA 21 1 showed a higher sensitivity in PE than in serum, probably because in a limited district tumor markers have concentrations higher than those in the blood stream. Therefore, the diagnosis of malignant PEs could be better achieved by measuring tumor markers in pleural fluid than in serum. The sensitivity of CEA was 30.6% in all malignant effusions and 52.5% in carcinomatous effusions. These percentages of sensitivity were lower than those described in some previous studies, 19,22,23 probably because of the prevalence in our series of patients with mesotheliomas over those with carcinomas. Only one of the mesothelioma patients (3.1%) showed CEA levels higher than the cutoff level. The cytologic analysis in a mesothelioma PE can be challenging even for a trained pathologist. Reviewing the literature, we could find only a few studies on tumor markers in PEs from mesothelioma, which were based on very small groups of patients 13,24 Because of a large number of mesothelioma cases in our series (32 patients), we were better able to define the diagnostic significance of CYFRA 21 1 and CEA in such a difficult diagnosis. CYFRA 21 1 sensitivity was significantly higher in patients with mesotheliomas than in those with carcinomas. In agreement with our results, other authors 24,25 have reported that mesothelioma cells expressed very high levels of cytokeratin 19 and that high levels of tissue polypeptide antigen (cytokeratins 8, 18, and 19) were present in the pleural fluid of patients with mesotheliomas. Salama et al, 24 in a series of nine mesothelioma patients, attributed a great clinical value to CYFRA 21 1 in the diagnosis of the disease. As reported by this author and others, 26,27 we found that high CYFRA 21 1 levels and low CEA levels are associated with mesothelioma. Because some lung carcinomas showed a similar behavior, the diagnosis of mesothelioma on the basis of this profile should be made with extreme caution. However, in our study, all the mesothelioma patients, except one, showed very low levels of CEA in all PEs. Consequently, we infer that high CYFRA 21 1 levels with low CEA levels are suggestive of mesothelioma, whereas high levels of CEA alone or high levels of both markers can show evidence of malignant PEs, excluding mesothelioma. As reported in the literature, 24 our findings indicated that CEA and CYFRA 21 1 are complementary tools to cytologic analysis in the clinical assessment of malignant PEs, in particular those resulting from mesothelioma. In our study, the cytologic CHEST / 119 / 4/ APRIL,

5 diagnostic sensitivity in mesothelioma PEs was 40.6%. In 19 mesotheliomas with a false-negative or uncertain cytology, the CYFRA 21 1 diagnostic sensitivity was 89%, while no cases showed CEA values higher than the cutoff level. The association of cytology with the measurement of PE CEA and CYFRA 21 1 levels allowed a correct diagnosis of 30 of 32 mesotheliomas (93.7%). In conclusion, when the cytologic analysis does not allow a final PE diagnosis, increased CYFRA 21 1 and/or CEA concentrations may represent, for the clinician, a useful decisional criterion before embarking on a more aggressive approach to treatment. When patients are in good performance status, pleural biopsy and/or thoracoscopy are necessary in order to stage the cancer and plan a correct therapeutic approach. In patients with poor clinical conditions, because of age or low performance status, diagnosis should be made on the basis of tumor markers alone, avoiding more aggressive diagnostic techniques. References 1 Millard FJC, Pepper JR. Pleural disease. In: Brewis RAL, Corrin B, Geddes DM, et al, eds. Respiratory medicine (vol 2). 2nd ed. London, UK: WB Saunders, 1995; Hausheer FH, Yarbro JW. Diagnosis and treatment of malignant pleural effusion. Semin Oncol 1985; 12: Loddenkemper R. Thoracoscopy-state of the art. Eur Respir J 1998; 11: Prakash UBS, Reiman HM. Comparison of needle biopsy with cytologic analysis for the evaluation of pleural effusion: analysis of 414 cases. Mayo Clin Proc 1985; 60: Light RW, Eozan YS, Ball WC. Cells in pleural fluid: their value in differential diagnosis. Arch Intern Med 1973; 132: Boutin C, Viallat JR, Cargnino P, et al. Thoracoscopy in malignant pleural effusions. Am Rev Respir Dis 1981; 124: Menzies R, Charbonneau M. Thoracoscopy for the diagnosis of pleural disease. Ann Intern Med 1991; 114: Plebani M, Basso D, Navaglia F, et al. Clinical evaluation of seven tumour markers in lung cancer diagnosis: can any combination improve the results? Br J Cancer 1995; 72: Wieskopf B, Demangeat C, Purohit A, et al. CYFRA 21 1 as a biologic marker of non-small cell lung cancer: evaluation of sensitivity, specificity, and prognosis. Chest 1995; 108: Cascinu S, Del Ferro E, Barbanti I, et al. Tumor markers in the diagnosis of malignant serous effusions. Am J Clin Oncol 1997; 20: Toumbis M, Rasidakis E, Passalidou E, et al. Evaluation of CYFRA 21 1 in malignant and benign pleural effusions. Anticancer Res 1996; 16: Villena V, Lopez-Encuentra A, Echave-Sustaeta J, et al. 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: Ferrer J, Villarino MA, Encabo G, et al. Diagnostic utility of CYFRA21 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: Peto J, Decarli A, La Vecchia C, et al. The European mesothelioma epidemic. Br J Cancer 1999; 79: Vercelli M, Orengo MA, Casella C, et al. Ligurian Tumor Registry, period ; In: Parkin DM, Whelan SL, Ferlay J, eds. Cancer incidence in five continents (vol II). Lyon, France: IARC Scientific Publications, Tamura S, Nishigaki T, Moriwaki Y, et al. Tumor markers in pleural effusion diagnosis. Cancer 1988; 61: Toumbis M, Chondros K, Ferderigos AS, et al. Clinical evaluation of four tumor markers in malignant and benign pleural effusions. Anticancer Res 1992; 12: San Jose ME, Alvarez D, Valdes L, et al. Utility of tumor markers in the diagnosis of neoplastic pleural effusion. J Clin Chim Acta 1997; 265: Shijubo N, Honda Y, Fujishima T, et al. Lung surfactant protein-a and carcinoembryogenic antigen in pleural effusions because of lung adenocarcinoma and malignant mesothelioma. Eur Respir J 1995; 8: Lai RS, Hsu HK, Lu JY, et al. CYFRA 21 1 enzyme-linked immunosorbent assay: evaluation as a tumor marker in nonsmall cell lung cancer. Chest 1996; 109: Satoh H, Sumi M, Yagyu H, et al. Clinical evaluation of CYFRA 21 1 in malignant pleural fluids. Oncology 1995; 52: Marel M, Stastny B, Melinova L, et al. Diagnosis of pleural effusions: experience with clinical studies, 1986 to Chest 1995; 107: Loncar R, Ostojic L, Tabakovic Loncar V, et al. Diagnostic potential of carcinoembryonic antigen and ferritine in tuberculous and malignant pleural effusion. Tumori 1995; 81: Salama G, Miedouge M, Rouzaud P, et al. Evaluation of pleural CYFRA 21 1 and carcinoembryonic antigen in the diagnosis of malignant pleural effusions. Br J Cancer 1998; 77: Bonfrer JMG, Schouwink JH, Kors CM, et al. CYFRA 21 1 and TPA as markers in malignant mesothelioma. Anticancer Res 1977; 17: Nisman B, Barak V, Heching N, et al. Cytokeratin markers in malignant pleural mesothelioma. Cancer Detect Prev 1998; 22: Mezger J, Lamerz R, Permanetter W. Diagnostic significance of carcinoembryonic antigen in the differential diagnosis of malignant mesothelioma. Thorac Cardiovasc Surg 1990; 100: Clinical Investigations